JAPAN INTERNATIONAL COOPERATION AGENCY (JICA)

MINISTRY OF TRANSPORT, SOCIALIST REPUBLIC OF VIETNAM (MOT)

HO CHI MINH CITY PEOPLE’S COMMITTEE (HCMC-PC)

THE STUDY ON
URBAN TRANSPORT MASTER PLAN
AND FEASIBILITY STUDY
IN HO CHI MINH METROPOLITAN AREA
(HOUTRANS)

FINAL REPORT

Volume 4: Transport Database

June 2004

ALMEC CORPORATION
 VOLUME 4: TRANSPORT DATABASE

TABLE OF CONTENTS

1. HOUTRANS TRANSPORT SURVEY AND DATABASE

1.1 Survey Description ................................................................................................ 1-1
1.2 Transport Database ............................................................................................... 1-3
2. INTEGRATION OF TRANSPORT AND GIS DATABASE
2.1 Background............................................................................................................ 2-1
2.2 GIS Data ................................................................................................................ 2-2
2.3 GIS Data and Transport Planning ......................................................................... 2-27
3. TRAFFIC DEMAND FORECAST USING CUBE/VOYAGER
3.1 Introduction ............................................................................................................ 3-1
3.2 2nd Stage Model in CUBE/Voyager ...................................................................... 3-2
3.3 3rd Stage Model..................................................................................................... 3-9
3.4 Traffic Assignment Result of 3rd Stage Model ...................................................... 3-27
4. TECHNOLOGY TRANSFER
4.1 Introduction ............................................................................................................ 4-1
4.2 Training Course 1 – HOUTRANS 2nd Stage Model ............................................. 4-1
4.3 Training Course 2 – HOUTRANS 2nd and 3rd Stage Models .............................. 4-3
4.4 Evaluation and Comments on Training Courses................................................... 4-3

APPENDICES

1 HOUTRANS Study Area Zone System Definition

2 CUBE/Voyager Model Description (1)
3 CUBE/Voyager Model Description (2)
4 HOUTRANS Training Course

i
 LIST OF TABLES

Table 1.1.1 Outline of Transport Surveys ....................................................................1-2

Table 1.2.1 HOUTRANS Transport Database (Primary and Secondary Data)...........1-5
Table 1.2.2 HOUTRANS Transport Database (Tertiary Data).....................................1-5
Table 2.2.1 HOUTRANS GIS Data ..............................................................................2-4
Table 2.2.2 HOUTRANS Data linked with GIS by Zone..............................................2-6
Table 2.2.3 HOUTRANS GIS Data and Renewal........................................................2-8
Table 2.2.4 Urban Planning and Efficient GIS Data ....................................................2-9
Table 2.3.1 Comparison between GIS and Demand Forecasting Software ...............2-28
Table 3.1.1 Key Changes of 3rd Stage Model Using CUBE/Voyager .........................3-1
Table 3.2.1 Calibrated Parameters for Trip Generation and Attraction Models ..........3-4
Table 3.2.2 Interzonal Trip Ratios by Area Types – Observed and Future .................3-5
Table 3.2.3 Calibrated Parameters of Interzonal Trip Distribution Model....................3-5
Table 3.2.4 Modal Split Model Generalized Cost Parameters by Mode .....................3-7
Table 3.2.5 Calibrated Modal Split Model Parameters ................................................3-7
Table 3.2.6 Base and Forecast Year Vehicle Occupancy and PCU Factors ..............3-8
Table 3.3.1 Trip Generation Model Parameters ..........................................................3-13
Table 3.3.2 Trip Attraction Model Parameters .............................................................3-13
Table 3.3.3 Trip Rate for Trip Generation Control Totals at Study Area Level ............3-14
Table 3.3.4 Interzonal Trip Ratios by Area Types – Observed and Future .................3-15
Table 3.3.5 Area Definition for the Application of Intrazonal Trip Proportions.............3-15
Table 3.3.6 HOUTRANS 3rd Stage Interzonal Trip Distribution Parameters ..............3-17
Table 3.3.7 Parameters for Walk Modal Split Model ...................................................3-19
Table 3.3.8 Modal Split Model Parameters..................................................................3-20
Table 3.3.9 Area Definition for the Application of Time Period Factors.......................3-24
Table 3.4.1 Assessment of M/P Network : PT 30% Case ...........................................3-27
Table 3.4.2 Assessment of M/P Network : PT 50% Case ...........................................3-28
Table 3.4.3 Comparison of Alternative Network Performance
based on 3rd Stage Model ....................................................................3-33
Table 4.4.1 Evaluation of the Training .........................................................................4-4
Table 4.4.2 Training Schedule .....................................................................................4-6

ii
 LIST OF FIGURES

Figure 2.2.1 Processing Flow of GIS Data....................................................................2-3

Figure 2.2.2 Hierarchy of GIS Data...............................................................................2-3
Figure 2.2.3 Application of GIS to Planning ..................................................................2-7
Figure 2.2.4 Illustration of GIS Data..............................................................................2-10
Figure 2.3.1 Estimate of Future Population Using GIS Data ........................................2-27
Figure 3.2.1 HOUTRANS Base Year 2nd Stage Model in CUBE/Voyager Software ...3-2
Figure 3.2.2 Comparison of Observed and Synthesized Trip Length Distribution .......3-6
Figure 3.2.3 Speed Flow Relationship Used for All Links in Assignment Model ..........3-8
Figure 3.3.1 HOUTRANS 3rd Stage Model – Overview of Model Configuration .........3-10
Figure 3.3.2 HOUTRANS 3rd Stage Model – Overview in CUBE/Voyager Platform ...3-11
Figure 3.3.3 Modal Split Model Process .......................................................................3-18
Figure 3.3.4 Trip Distribution by Time of Day by Trip Purpose .....................................3-22
Figure 3.3.5 % of Trip by Time of Day by Trip Purpose ................................................3-22
Figure 3.3.6 Trip Distribution by Time of Day by Mode of Travel..................................3-23
Figure 3.3.7 % of Trip by Time of Day by Mode of Travel.............................................3-23
Figure 3.3.8 Speed Flow Relationship for 3rd Stage Assignment Model......................3-26
Figure 3.4.1 Traffic Assignment on Recommended Network : PT 30% Case
at AM Peak Hour ...................................................................................3-29
Figure 3.4.2 Traffic Assignment on Recommended Network : PT 30% Case
at Off Peak Hour....................................................................................3-30
Figure 3.4.2 Traffic Assignment on Recommended Network : PT 50% Case
at AM Peak Hour ...................................................................................3-31
Figure 3.4.2 Traffic Assignment on Recommended Network : PT 50% Case
at Off Peak Hour....................................................................................3-32

iii
 The Study on Urban Transport Master Plan and Feasibility Study in HCM Metropolitan Area (HOUTRANS)
Final Report
VOLUME 4: Transport Database

1 HOUTRANS TRANSPORT SURVEY AND DATABASE

1.1 Survey Description

Transport and traffic surveys were conducted to complete the Transport Master Plan and
feasibility studies as shown in Table 1.1.1. They were done in order to acquire sufficient
information, which were not available from existing sources. The HIS is one of the most
important data, it surveyed about 28,000 households (1.8% of the whole households) in
the study area. Daily trip information, socio-economic status of residents in the study area
as well as opinions and problems, the residents were encountering, was obtained through
the HIS. The result of the household interview can be a very important database not only
for traffic demand forecasting but also for social and economical aspects since the sample
size is big enough to accommodate various sectional analyses such as income levels,
residential areas, car ownership, and so on.

Other surveys were conducted to compensate information from the HIS. Traffic count data
can be used for adjusting the results of traffic demand forecasting in order to get more
accurate forecasting results. These data can be also used as a benchmark to check how
the traffic volumes and vehicle proportions change by comparing time series data.
Therefore, these results are also very important in terms of the component of transport
database.

Detail information regarding the HIS, and the transport and traffic surveys are described in
Technical Report No.1 and No.2 respectively.

1-1
 The Study on Urban Transport Master Plan and Feasibility Study in HCM Metropolitan Area (HOUTRANS)
Final Report
VOLUME 4: Transport Database
Table 1.1.1 Outline of Transport Surveys

Survey Objectives Coverage Method

1. Person Trip • Socio-economic profile and trip • 28,000 sample households (1.8% • Direct interview with
Survey information of residents of the total number of households household head and
(Household • Perception of residents on in HCMC except Cu Chi and Can members
Interview transport services and Gio, and 0.9% of ones in the
Survey) environment adjoining area)
2. Cordon Line • Traffic volume on cordon line • 22 stations on the study area • Direct interview with
Survey (vehicle/passenger) boundary drivers/passengers at
• Socio-economic profile and trip • 4 stations at transport terminals roadside
information of residents outside (airport, inland ferry and railway • Traffic count (vehicle)
the study area station) • 24 and 18 hours
3. Screen Line • Traffic volume on screen line • 37 stations on 2 east-west • Traffic count
Survey (vehicle/passenger) screen lines (vehicle/passenger)
• 24 and 18 hours
4. Traffic Count • Traffic volume at major road • 16 stations • Traffic count (vehicle)
Survey sections • 24 and 18 hours
5. Intersection • Traffic volume at congested • 12 major intersections • Traffic count (vehicle)
Traffic Count intersections by turning direction • 12 hours (6-18)
Survey
6. Travel Speed • Travel speed on major routes by • 15 routes • Floating car method
Survey section • Car, motorcycle, bicycle, • 6 round trips by route
• Bottleneck sections bus/lambro by vehicle type
• 12 hours
7. Public Transport • Travel characteristics • Bus/lambro, cyclo, taxi and bike • Direct interview with
Passenger • Perception of users on existing taxi public transport
Interview Survey and proposed public transport • About 2,000 passengers passengers
services • 12 hours
8. Public Transport • Capacity and utilization of • 11 major bus/lambro terminals • Facility inventory
Terminal Survey terminals survey
• Characteristics of bus/lambro • Counting frequencies
operation and passengers by
route
• 18 hours
9. Public Transport • Boarding and alighting • 19 representative routes • On-board observation
On-board passengers by route section • 13 hours • 6 round trips by route
Survey
10. Public • Operational and financial • 30 bus/lambro operators • Direct interview
Transport characteristics of public transport • About 300 drivers of cyclo, taxi
Operator Survey operators and bike taxi
11.Private Mode • Travel route choice • Private mode users (car, • Direct interview with
User Interview • Perception on road services and motorcycle, bicycle) users and students
Survey willingness- to-use of public • Students • 12 hours
transport • About 2,000 users
12. Parking Survey • Inventory at City Center • All parking facilities at Dist.1, 3 • Facility inventory
• Parking capacity and utilization and 5 survey
• Perception of users on parking • 20 sites: on-/off-street parking for • Utilization survey
services car and motorcycle/bicycles • Direct interview with
users
• 18 hours
13.Freight • Characteristics of freight vehicle • About 90 transport operators • Direct interview with
Transport transport • 17 sites of Ports, Airport, ICDs, operators
Survey for traffic count and OD interview • Traffic count and OD
interview at gates
• 24 hours
14.Establishment • Opinions and assessment from • About 800 establishments: 200 • Direct interview with
Survey offices, schools, hospitals, etc., for educational, 75 for medical , establish managers
on present transportation 250 for commercial, 110 for
situation. office, 125 for logistics, 20 for
urban service, 20 for
recreational facilities.
15.Inland • Traffic volume of inland waterway • 6 major waterways (rivers & • Traffic Count and OD
Waterway freight traffic creeks) interview at major
Transport • Characteristics of inland • 14 hours points in waterway
Survey waterway freight traffic and freight • 10 freight operators • Direct interview with
operators freight operating
companies
Source: Study Team

1-2
 The Study on Urban Transport Master Plan and Feasibility Study in HCM Metropolitan Area (HOUTRANS)
Final Report
VOLUME 4: Transport Database
1.2 Transport Database
1) Introduction
As briefly stated in the previous section, 15 kinds of transport surveys were conducted for
the HOUTRANS. The results of the surveys were compiled in the form of a database
together with other relevant data that were independently prepared by government
agencies. It is very useful to develop such a comprehensive database for general use in
transport planning, policy-making, management, and project appraisals in the field of
urban transportation and development. The database was fully utilized in the course of the
analysis and planning stages of the HOUTRANS itself and is expected to be used in other
succeeding studies and planning after the study’s completion. The HOUTRANS database
was developed based on the following policies:

1) It should have a simple and clear structure for easy understanding of users.
2) It should have a program package to compile a requested data file for easy data
search and retrieval.

3) It should allow data output in the form of hard copy, map or file in a data storage
media such as CD-ROM and so on.

4) It should be readable with prevailing software for database handling, map drawing
and transportation planning.

5) It should be directly accessible to users through a computer network.

2) Coverage of Database
The HOUTRANS transport database consists of major results from the field and interview
surveys conducted for the study and other related data, such as population, road
inventory, etc., obtained from other sources. The data are classified into seven categories
as listed below:

1) Socio-economic data
a) Population and household
b) Employment by workplace

c) Workers by residence

d) School attendance

e) Income level
f) Vehicle ownership

2) Transport demand

a) OD matrices

b) Other demand data

1-3
 The Study on Urban Transport Master Plan and Feasibility Study in HCM Metropolitan Area (HOUTRANS)
Final Report
VOLUME 4: Transport Database
3) Road and traffic data

a) Road network

b) Road inventory
c) Traffic volume count

d) Travel speed

4) Public transportation data

a) Bus route map and operational characteristics

b) Terminal location map and data

c) Passenger/driver/operator interview survey data

5) Other transport related data

a) Port/airport traffic

b) Water transport-related data

c) Parking data

d) Truck survey data

e) Traffic accident data

f) Establishment data

g) Willingness-to-pay survey data

3) Data Category
The data can be categorized from the viewpoint of data processing. The first category is a
group of unprocessed data such as the traffic-count data and the HIS master file. These
raw data will be useful for a user who has particular interest in analysis.
The second category is a group of processed data such as various kinds of OD matrices
and a computerized transport network for simulation work. Many tables, maps, charts, and
graphs are included in the database for general use. Both categories are related to the
current simulation. Therefore, these data can be used by any party even as mere facts, if
no specific accuracy is required.
The third category is a group of forecast data such as socio-economic parameters, OD
matrices and road network. All of them are scenario-specific, i.e. they were projected
based on a number of assumptions according to the selected scenario for urban and
transport development. They are not mere facts, and, therefore, require careful handling
by database users. When these data are open to external users, database managers
must take necessary measures to avoid misunderstanding and misuse. Table 1.2.1 and
1.2.2 summarize the primary and secondary, and tertiary data, respectively.

1-4
 The Study on Urban Transport Master Plan and Feasibility Study in HCM Metropolitan Area (HOUTRANS)
Final Report
VOLUME 4: Transport Database
Table 1.2.1 HOUTRANS Transport Database (Primary and Secondary Data)

Primary Data Secondary Data

Data Category
(Original Data) (Processed Data)
• HIS Master File • Population Data
• Population/Household • Employment
Socio-economic • Zone Map • School Attendance
• Income
• Vehicle Ownership
• HIS Master File • OD Trip Data
Transport Demand
• Demand Data
• Road Inventory • Road Network
• Traffic Count • Road Facilities
Road and Road Traffic
• Travel Speed • Road Traffic
• Intersection Traffic
• Bus Data • Bus routes and
• Public Transport Terminal operation
Public Transport Data characteristics
• Passenger/driver/operator • Terminal Location
Interview Survey Data
• Port/airport Traffic Data
• Inland Waterway
Transport
Other Transport Related • Parking Data
• Traffic Accident Data
• Establishment Data
• Willingness-to-pay Data
Source: Study Team

Table 1.2.2 HOUTRANS Transport Database (Tertiary Data)

Data Category Year

• Socio-economic Parameters (Population, employment,

income, etc.) 2010, 2020

• Road and Transit Network Data 2010, 2020

• OD Matrices – person trip basis 2010, 2020

Source: Study Team

1-5
 The Study on Urban Transport Master Plan and Feasibility Study in HCM Metropolitan Area (HOUTRANS)
Final Report
VOLUME 4: Transport Database

2 INTEGRATION OF TRANSPORT AND GIS DATABASE

2.1 Background
The efficient usage of GIS (geographical information system) in transport planning is very
significant from the viewpoint of spatial understanding and evaluation of road and other
transport facility development as well as time-series and cross sectional data handling and
management by using it as a data platform. When GIS is used as a data platform for
transport planning, five types of data can be included: a) Natural Condition, b) Urban
Infrastructure, c) Land Use, d) Transport Facilities, and e) Traffic Demand. From a) to d),
the data can be managed as a static data, dot, line or polygon. On the other hand, data e)
should be handled as a drastic data that change by time.

These data are closely related to the evaluation methodology of the planning. From a) to
c), these data are crucial elements for analyzing and estimating existing and future traffic
demand and these data have played a significant role in transport planning. However,
these data can become easier, more understandable and subjective with the use of the
GIS.
In addition to this, traffic information has been managed by traffic zone when the demand
forecasting is conducted in the four-step method. In other words, “dot information”,
individual information, has been converted to “area information” aggregated by zone.
Recently, disaggregated behavioral model and other individual traffic behavior model have
been developed well in the transportation research and GIS can play as a platform for
handling such individual data, “dot and line” data.
In HOUTRANS, in the process of traffic demand forecasting, GIS data were used
sufficiently as a platform of database while the conventional four-step method was
adopted.

2-1
 The Study on Urban Transport Master Plan and Feasibility Study in HCM Metropolitan Area (HOUTRANS)
Final Report
VOLUME 4: Transport Database
2.2 GIS Data
1) Outline
GIS data relating to land use are classified into three parts: “Land use”, “Socio-economic”,
“Transport”. “Land use” includes basic information, such as boundary, natural condition,
land condition, land use and existing master plan. These are useful in planning
land-based planning. “Socio-Economic” includes basic information, such as urban
infrastructure, urban facility and social indices, are useful in analyzing area characteristics.
“Transport” includes basic information, such as existing transport facility, ongoing and
committed projects and related facilities, which were classified as transportation facility
information. These data are shown in Table1. They were transformed into digital data
based on existing maps and developed into a GIS database by linking information such as
name, category, area, and so on.

2) Data Processing
The data was collected as follows;

- Basic geographical information (boundaries, rivers, etc.) : gotten from commercial

data;
- Information relating to land use: gotten from UPI (Ho Chi Minh City)
- Data about surrounding provinces: NAGECCO working for consultation to master
plans of provinces.
- Information relating to transportation: TUPWS

To process these data, we used GIS application and transformed the CAD data to GIS
data.
The gathered data were transformed through the following steps;
- Declaring name and address.

- Clarifying scale by measuring the area and length.

Classifying categories to develop hierarchical data structure and inputting the data. For
example, facility data can be categorized by sort (school, hospital, etc) and type (ex.
School, primary and secondary school, high school, university, and so on)

2-2
 The Study on Urban Transport Master Plan and Feasibility Study in HCM Metropolitan Area (HOUTRANS)
Final Report
VOLUME 4: Transport Database
Figure 2.2.1 Processing Flow of GIS Data

Original Data

Input relating information Following Study

- Data with name or address - in case gathered
- Data with area or length data has not
- Category enough information,
following study
must be conducted.

GIS Data

Source: Study Team

The database developed as GIS can be classified into hierarchies and they are shown in
Figure 2.2.2. It is possible to analyze deeper. It becomes possible to analyze more
profound by extracting and combining necessary information. However, linking
socio-economic indices with geographical data make it possible for spatial analysis such
as population distribution, jobless distribution by district (or zone), and so on.
Figure 2.2.2 Hierarchy of GIS Data

* Line data & point data such as

Socio-Economic urban infrastructure, urban
facility, social condition, etc.

* Line data such as road, railway,

Transport etc.

* Area data such as landuse, land

Landuse condition, etc.

* Area data & line data such as

Geographical Data (=Base map) city, district, province, zone,
etc.
* Area data such as river, contour
line, building, etc.
Source: Study Team

2-3
 The Study on Urban Transport Master Plan and Feasibility Study in HCM Metropolitan Area (HOUTRANS)
Final Report
VOLUME 4: Transport Database

Table 2.2.1 HOUTRANS GIS Data

[Base map]
Category Index Coverage Data Source Year
1.Boundary Prefecture Boundary All Study Area DolSoft GIS data
District Boundary (Private Company)
Ward Boundary
Zone ST
2.Natural River/Canal All Study Area DolSoft GIS data
Condition Contour Line HCMC (not complete)

[Land use]
Category Index Coverage Data Source Year
3.Land Condition Flood Prone Area by Rain, All Study Area JICA SFEZ Study 1999
Flood Prone Area by Tide NAGECCO 1995
Good Soil Geological Map 2003

4.Landuse Province HCMC, UPI 2000

Long An NAGECCO 2003
District Di An, Thu Dau Mot, NAGECCO 2003
Thuan An, Bien Hoa,
Long Thanh, Tan An
5.Existing Province HCMC in 2020, UPI 1998
Master Plan Long An in 2020 NAGECCO
District Di An in 2010 NAGECCO 2003
Thu Dau Mot in 2020
Thuan An in 2020
Bien Hoa in 2020
Long Thanh in 2020
Tan An in 2020
Source: Study Team

2-4
 The Study on Urban Transport Master Plan and Feasibility Study in HCM Metropolitan Area (HOUTRANS)
Final Report
VOLUME 4: Transport Database
Table 2.2.1 HOUTRANS GIS Data (continued)

[Socio-economy]
Category Index Coverage Data Source Year
1)
6.Urban Water Supply HCM urbanized area, JICA 1999
Infrastructure Bien Hoa, Di An, Nhon NAGECCO 2003
Trach, Tan An, Thu Dau
Mot
1)
Drainage HCM urbanized area, JICA 1999
Bien Hoa, Di An, Tan
An, Thu Dau Mot
Solid Waste (Route & Pick HCMC ST 2003
-up Point & Treatment Site)
Rubbish Collection Area 2003
Bien Hoa, Nhon Trach, NAGECCO
Tan An, Thu Dau Mot,
Thuan An
7. Establishment School HCMC ST 2003
Bien Hoa, Di An, Nhon NAGECCO 2003
Trach, Tan An, Thu Dau
Mot, Thuan An
Hospital HCMC ST 2003
Bien Hoa, Di An, Nhon NAGECCO 2003
Trach, Tan An, Thu Dau
Mot, Thuan An
Hotel HCM urbanized area ST 2003
Bien Hoa, Di An, Thu NAGECCO 2003
Dau Mot
Market HCMC ST 2003
Bien Hoa, Di An, Nhon NAGECCO 2003
Trach, Tan An, Thu Dau
Mot, Thuan An
8.Social Low Income Area HCM urbanized area NRED 2002
Information (Temporary residence area) Bien Hoa, Di An, Nhon NAGECCO 2003
Trach, Tan An, Thu Dau
Mot, Thuan An
Historical Building HCM urbanized area Dr. Ohta 1996
1) The Study on Urban Drainage and Sewerage System for HCMC (JICA, 1999)

[Transport]
Category Index Coverage Data Source Year
9.Existing Railway(VNR) All Study Area 2003
Infrastructure Port/Airport
Waterway
10.Project Intersection HCM urbanized area
Signal
Road TUPWS
Parking
11.Bus Bus Stop HCMC ST
Operation 2003
12.Accident Accident Prone Location 2003
Source: Study Team

2-5
 The Study on Urban Transport Master Plan and Feasibility Study in HCM Metropolitan Area (HOUTRANS)
Final Report
VOLUME 4: Transport Database
Table 2.2.2 HOUTRANS Data linked with GIS by Zone

Category Index Coverage Data Source Year

Area Total, river, land excluding All Study Area ST
river
Population Population, Population density HCMC except Cu Chi, MVA 1996,
Can Gio District
All Study Area ST 2002,
2020
Population growth rate HCMC except Cu Chi, ST 96-02,
Can Gio District
All Study Area 02-20
No. of household All Study Area ST 2002
Employment No. of employment at All Study Area ST 2002,
residence/workplace and 2020
jobless rate
Student No. of students at All Study Area ST 2002,
residence/school place 2020
Day / night ratio Total/employment/student All Study Area ST 2002,
2020
Source: Study Team

3) Application of GIS to Planning Process

Spatial analysis using the GIS data contributes to the planning methods especially on
spatial planning such as land-use distribution planning, industrial park allocation planning,
and so on.
A plan should be developed based on various information which are applicable to objective
of the plan. This diverse information should be organized genealogically, and one method
that expresses the data visually is GIS. By using this system, it allows not only area-based
analysis but also spatial analysis by considering quantitative data. It provides much useful
information to solve problems raised on planning processes. Since GIS can manage
multifarious data into a unitary whole and keep basic information of urban planning its merit
is to provide coherence between different plans. The figure below shows the flow of
applying GIS to planning.

2-6
 The Study on Urban Transport Master Plan and Feasibility Study in HCM Metropolitan Area (HOUTRANS)
Final Report
VOLUME 4: Transport Database
Figure 2.2.3 Application of GIS to Planning

Plan/Project
- Land-use Planning
GIS Data Other Information - Industrial Planning
- Map - Economic - Agricultural Planning
- Population - Finance - Urban Planning
- Land use - Social Issues - Transportation Network
etc. etc. Planning
etc.

Analysis
Problem-finding
Countermeasures
Policy Developing
Planning

Source: Study Team

4) Issues for developing and managing GIS database

(1) Maintenance and renewal

- Once the database is developed, it should be managed by proper personnel or an

organization with knowledge on the GIS in order to maintain the latest data. And it should
be prepared in an environment that provides easy access and referencing for planning
specialists (See Table 2.2.2).
- Data renewal within appropriate period by conducting studies regularly is necessary.

(2) Collection of new data and development of database

- The data which the Study Team collected is the most basic data relating to land use and
urban planning. It is necessary to develop more comprehensive database according to
various needs for urban planning.
(3) Efficient data

- It can be considered that the data shown in Table 2.2.3 will be needed depending on the
objective of each planning.

2-7
 The Study on Urban Transport Master Plan and Feasibility Study in HCM Metropolitan Area (HOUTRANS)
Final Report
VOLUME 4: Transport Database
Table 2.2.3 HOUTRANS GIS Data and Renewal

Period of Development Action5)

Timing of Data Object on
Data Category Index Newness1) Scale3)
Renewal2) Unit4) Data GIS6)
Survey ＤＢ ＧＩＳ
Coding

Provincial Boundary Ｌ Ｓ Ｌ Ｓ Ｍ Ｒ
Admin. District Boundary Ｌ Ｓ Ｌ Ｓ Ｍ Ｒ
Boundary Ward Boundary Ｌ Ｓ Ｌ Ｓ Ｍ Ｒ
Zone Boundary Ｌ Ｓ Ｌ Ｌ Ｒ
(1)Natural
Topographic Map Ｌ Ｌ Ｌ Ｌ Ｌ
Condition
Contour Map Ｌ Ｌ Ｌ Ｓ Ｌ Ｒ
Natural
River Ｌ Ｌ Ｌ Ｍ Ｒ
Condition
Flooded Area Ｒ
Soil Condition Ｌ Ｓ Ｌ Ｒ
School Ｌ Ｍ Individual Ｌ Ｓ Ｍ Ｓ Ｐ
Hospital Ｌ Ｓ Individual Ｍ Ｓ Ｍ Ｓ Ｐ
Hotel ○ Ｍ Ｍ Individual Ｌ Ｓ Ｍ Ｓ Ｐ
Restaurant Ｓ Ｌ Individual Ｌ Ｍ Ｌ Ｍ Ｐ
Market ○ Ｍ Ｍ Individual Ｌ Ｓ Ｍ Ｓ Ｐ
Infrastructure Industrial Park Ｌ Ｓ Individual Ｓ Ｓ Ｍ Ｓ Ｒ
(2) Infra- Water Supply Coverage Ｌ Ｐ、Ｎ
structure
Drainage Coverage Ｌ Ｎ
Solid Waste Correction
Ｍ Ｒ
Coverage
Solid Waste Disposal
Ｌ Ｐ
Facility
Social Low Income Area Ｍ Ｒ
Environment Historical Building Ｌ Ｐ
Population Ｍ Ｌ Object Ｌ ＊
Existing
(3) Land Use Land Use ○ Ｌ Ｌ Ｌ Ｒ／Ｂ
Planning Land Use Plan ○ Ｒ
Road Network Ｍ Ｌ Ｌ Ｍ Ｌ Ｌ Ｎ
Railway Network Ｍ Ｍ Ｍ Ｓ Ｍ Ｓ Ｎ
Bus Network Ｓ Ｌ Ｌ Ｍ Ｌ Ｌ Ｎ
Existing Airport Ｌ Ｓ Ｓ Ｓ Ｓ Ｓ Ｐ／Ｒ
Port Ｌ Ｓ Ｍ Ｓ Ｓ Ｓ Ｐ
(4)Transport
Inland waterway Ｌ Ｍ Ｌ Ｍ Ｍ Ｍ Ｎ
Facilities
Parking Ｍ Ｌ Ｌ Ｍ Ｌ Ｌ Ｐ
Intersection Ｐ
Signal Ｐ
Planning
Road Network Ｌ
Parking Ｐ
Traffic Volume Ｓ Ｌ Object Ｌ Ｍ ＊
(5)Traffic Demand No. of Passenger by
Ｓ Ｌ Object Ｌ Ｍ ＊
Route
Source: Study Team
1) : Very important. ○: Important
2) L: Long, M: Medium, S: Short
3) L: Large, M: Medium, S: Small
4) Individual: Facilities have attributes, Object: Other objects have attributes
5) L: Long, M: Medium, S: Short
6) P: Plot, L: Line, N: Network, B: Block, R: Region

2-8
 The Study on Urban Transport Master Plan and Feasibility Study in HCM Metropolitan Area (HOUTRANS)
Final Report
VOLUME 4: Transport Database
Table 2.2.4 Urban Planning and Efficient GIS Data

Plan Data Common data

Prospect and distribution of population Socio-economic indices

(framework)

Industrial zoning plan (Agricultural Purpose of building,

preservation, Industrial park arrangement, population density,
commercial complex etc.) working/non-working
population distribution

Water supply plan River, existing pipe line,

agricultural plan, industrial
zone

Power supply plan Existing power line, power Population, boundary,

station, industrial park and urban infrastructure,
housing allocation land use, road network,
etc.
Road development plan (planning, Traffic volume, public
maintenance, etc) transport plan

Public transport node allocation plan Public transport plan, traffic

volume

Garbage collection plan Housing allocation, volume of

garbage, garbage collection
plan, garbage disposal facility
allocation

Park development and arrangement plan Existing park allocation

Source: Study Team

2-9
 The Study on Urban Transport Master Plan and Feasibility Study in HCM Metropolitan Area (HOUTRANS)
Final Report
VOLUME 4: Transport Database
Figure 2.2.4 Illustration of GIS Data

1. Boundaries

2. Natural Conditions

Source: Study Team

2-10
 The Study on Urban Transport Master Plan and Feasibility Study in HCM Metropolitan Area (HOUTRANS)
Final Report
VOLUME 4: Transport Database
Figure 2.2.4 Illustration of GIS Data (Continued)

3. Land Conditions

4. Land Uses

Source: Study Team

2-11
 The Study on Urban Transport Master Plan and Feasibility Study in HCM Metropolitan Area (HOUTRANS)
Final Report
VOLUME 4: Transport Database
Figure 2.2.4 Illustration of GIS Data (Continued)

5. Master Plan

6-1. Urban Infrastructure (Water Supply)

Source: Study Team

2-12
 The Study on Urban Transport Master Plan and Feasibility Study in HCM Metropolitan Area (HOUTRANS)
Final Report
VOLUME 4: Transport Database
Figure 2.2.4 Illustration of GIS Data (Continued)

6-2. Urban Infrastructure (Drainage)

6-3. Urban Infrastructure (Drainage Zone and Flood-prone Area in HCMC)

Source: Study Team

2-13
 The Study on Urban Transport Master Plan and Feasibility Study in HCM Metropolitan Area (HOUTRANS)
Final Report
VOLUME 4: Transport Database
Figure 2.2.4 Illustration of GIS Data (Continued)

6-4. Urban Infrastructure (Solid Waste Collection Area and Route/Treatment Site in HCMC)

6-5. Urban Infrastructure (Solid Waste Collection Area in Surrounding Provinces)

Source: Study Team

2-14
 The Study on Urban Transport Master Plan and Feasibility Study in HCM Metropolitan Area (HOUTRANS)
Final Report
VOLUME 4: Transport Database
Figure 2.2.4 Illustration of GIS Data (Continued)

7. Establishments (Schools, Hospitals, Hotels, and Markets)

8-1. Social Information (Low-income Areas in HCM Urbanized Area)

Source: Study Team

2-15
 The Study on Urban Transport Master Plan and Feasibility Study in HCM Metropolitan Area (HOUTRANS)
Final Report
VOLUME 4: Transport Database
Figure 2.2.4 Illustration of GIS Data (Continued)

8-2. Social Information (Low-income Areas in Surrounding Provinces)

Bien Hoa

Nhon Trach

Di An Tan An

Thu Dau Mot

Thuan An

Source: Study Team

2-16
 The Study on Urban Transport Master Plan and Feasibility Study in HCM Metropolitan Area (HOUTRANS)
Final Report
VOLUME 4: Transport Database
Figure 2.2.4 Illustration of GIS Data (Continued)

8-3. Social Information (Historical Buildings)

[Transport]
9-1. Existing Infrastructure (Facilities)

Source: Study Team

2-17
 The Study on Urban Transport Master Plan and Feasibility Study in HCM Metropolitan Area (HOUTRANS)
Final Report
VOLUME 4: Transport Database
Figure 2.2.4 Illustration of GIS Data (Continued)

9-2. Existing Infrastructure (Intersections)

9-3. Existing Infrastructure (Signals)

Source: Study Team

2-18
 The Study on Urban Transport Master Plan and Feasibility Study in HCM Metropolitan Area (HOUTRANS)
Final Report
VOLUME 4: Transport Database
Figure 2.2.4 Illustration of GIS Data (Continued)

9-4. Existing Infrastructure (Parking Areas)

Location of Parking
by Ow nertype
District
Compary/Agency
Building ow ner
House ow ner
Others

0 0.5 1
kilometers

10. Road Projects

Source: Study Team

2-19
 The Study on Urban Transport Master Plan and Feasibility Study in HCM Metropolitan Area (HOUTRANS)
Final Report
VOLUME 4: Transport Database
Figure 2.2.4 Illustration of GIS Data (Continued)

11. Bus Operation (Bus Stops)

12. Accident-prone Locations

Source: Study Team

2-20
 The Study on Urban Transport Master Plan and Feasibility Study in HCM Metropolitan Area (HOUTRANS)
Final Report
VOLUME 4: Transport Database
Figure 2.2.4 Illustration of GIS Data (Continued)

13. Population, 2002

14. Population Density, 2002

Source: Study Team

2-21
 The Study on Urban Transport Master Plan and Feasibility Study in HCM Metropolitan Area (HOUTRANS)
Final Report
VOLUME 4: Transport Database
Figure 2.2.4 Illustration of GIS Data (Continued)

15. Households, 2002

16. Population Growth, 2002-2020

Source: Study Team

2-22
 The Study on Urban Transport Master Plan and Feasibility Study in HCM Metropolitan Area (HOUTRANS)
Final Report
VOLUME 4: Transport Database
Figure 2.2.4 Illustration of GIS Data (Continued)

17. Employment at Residence, 2002

18. Employment at Workplace, 2002

Source: Study Team

2-23
 The Study on Urban Transport Master Plan and Feasibility Study in HCM Metropolitan Area (HOUTRANS)
Final Report
VOLUME 4: Transport Database
Figure 2.2.4 Illustration of GIS Data (Continued)

19. Students at Residence, 2002

20. Students at School Place, 2002

Source: Study Team

2-24
 The Study on Urban Transport Master Plan and Feasibility Study in HCM Metropolitan Area (HOUTRANS)
Final Report
VOLUME 4: Transport Database
Figure 2.2.4 Illustration of GIS Data (Continued)

21. Unemployment Rate, 2002

22. Day / Night Ratio (Total), 2002

Source: Study Team

2-25
 The Study on Urban Transport Master Plan and Feasibility Study in HCM Metropolitan Area (HOUTRANS)
Final Report
VOLUME 4: Transport Database
Figure 2.2.4 Illustration of GIS Data (Continued)

23. Day / Night Ratio (Employment), 2002

24. Day / Night Ratio (Students), 2002

Source: Statistical Yearbook, HIS, Estimated by Study Team

2-26
 The Study on Urban Transport Master Plan and Feasibility Study in HCM Metropolitan Area (HOUTRANS)
Final Report
VOLUME 4: Transport Database
2.3 GIS Data and Transport Planning
1) Land Use and Traffic Demand
As described in the previous section, GIS is very effective in urban planning (land-use
planning) for understanding existing urbanized situations, good condition areas to develop,
and so on. In understanding existing urbanized situations, population density is one of the
key indices to grasp the stage of urbanization. In the HOUTRANS, habitable areas, which
excluded lands for transport facilities and rivers, by traffic zone were calculated using GIS
data and software.

For analyzing land potentials for further development, various forms of information were
examined using the GIS software by each zone as shown below:

- Ratio of built-up and nonbuilt-up areas.

- Ratio of high-density, middle-density and low-density in the builtup areas.

- Ratio of inundated problem areas and low-lying or hilly areas.

- Estimated access time to the CBD.

- Development level of infrastructure (water supply, drainage, Solid Waste Correction

Coverage, and so on)
- Access to urban services (market, hospital, school, and so on)

- Traffic condition (road density, access to the primary roads, congestion, and so on)

These information and population density in habitable area, scientific and logical land use
planning can be conducted and it makes the traffic demand forecasting more likely.
Figure 2.3.1 Estimate of Future Population Using GIS Data

Natural Condition
Infrastructure
Land Use

Potential for Population Density

Development in Habitable Area

Selection of
Suitable Area for
Development

Future Land Use

Future Day/ Night

Population by Zone

Source: Study Team

2-27
 The Study on Urban Transport Master Plan and Feasibility Study in HCM Metropolitan Area (HOUTRANS)
Final Report
VOLUME 4: Transport Database
2) Road Network Database
One of the most important databases in transport planning is road network. Road network
database can be used not only as a visual or spatial understanding of present conditions,
but also as a basis for network analysis in traffic demand forecasting. Generally, the type of
network database depends on the demand forecasting software. The JICA STRADA and
the CUBE/Voyager were used in this study and these are specialized as traffic demand
forecasting software not as GIS software.

The comparison between GIS software and traffic demand forecasting software in transport
planning and database is illustrated in Table 2.3.1. GIS software has a simple design and
can be used by everybody while traffic demand forecasting is good for experts. Demand
forecasting software is more analytical and the function of spatial analysis is limited and
rather static while GIS software is flexible and variable for data handling. The purposes of
these software are different so that for better analysis, organic connection between GIS
and traffic demand software is required.

In HOUTRANS, these two systems were developed for better connection between them:
- Conversion system from GIS road data to JICA STARA link data: road information in GIS
database (MapInfo format) can be converted to JICA STRADA format by using a
computer program developed by the study team. This can help to store various
attributes of road on MapInfo format and data handling is very simple.

- Calculation system for QV: The setting for road QV depended on the transport experts.
The calculation system of QV using MapInfo was invented by the Study Team and
non-experts can acquire road QV by inputting simple attributes like lane width, class of
road and so on.
Table 2.3.1 Comparison between GIS and Demand Forecasting Software

GIS Traffic Demand Forecasting

Software MapInfo, Arc Info, etc JICA STRADA, CUBE/Voyager, etc.
User General users Experts
Function - Find/Categorization - Forecasting/Calculation
- Visual - Analytical
- Easy Management for Spatial - Capacity of Spatial Information is
Information limited
- Spatial Analysis - Quantitative Analysis
Data Structure - Administrative/Planning Base - Analysis Base
- Unit: Road - Unit: Link
- Wider Information and - Limited Information and Structure
Flexible Structure - Planning Indices: QV, Capacity,
- Planning Indices: Lane width, etc.
Class of Road, etc.
Source: Study Team

2-28
 The Study on Urban Transport Master Plan and Feasibility Study in HCM Metropolitan Area (HOUTRANS)
Final Report
VOLUME 4: Transport Database

3 TRAFFIC DEMAND FORECAST USING CUBE/VOYAGER

3.1 Introduction
Traffic demand forecasting for the HOUTRANS was done by using two kinds of computing
software as it was described in the last chapter: JICA STRADA and CUBE/Voyager. As
explained in the main text, there are three stages on HOUTRANS traffic demand
forecasting systems and initially, JICA STRADA was used for the 1st and 2nd stage models
and CUBE/Voyager was introduced to the 2nd and 3rd stage models specifically. Here, the
detailed information on CUBE/Voyager of the 2nd and 3rd model are described in the light
of actual computing program on demand forecasting.
rd
Table 3.1.1 Key Changes of 3 Stage Model Using CUBE/Voyager
nd rd
2 Stage Model 3 Stage Model
Trip ・ Input: population at ・ Input: population, number of
Generation/Attraction resident, number of workers and number of
workers at workplace, students at resident, number of
number of students at workers at workplace, number
school of students at school
・ Trip Purpose: none ・ Trip Purpose: 5 types (to work,
・ Household Vehicle to school. private, business, to
Ownership: none home)
・ Personal Vehicle ・ Household Vehicle Ownership:
Ownership: none 4 types (car owner, multi M/C
owner, single M/C owner, non
motorized vehicle owner)
・ Personal Vehicle Ownership: 2
types (owner, non owner)
Trip Distribution ・ Trip Purpose: none ・ Trip Purpose: 5 types (to work,
・ Household Vehicle to school. private, business, to
Ownership: none home)
・ Personal Vehicle ・ Household Vehicle Ownership:
Ownership: none 4 types (car owner, multi M/C
・ Assigned traffic on the owner, single M/C owner, non
OD pairs which have motorized vehicle owner)
traffic in base year ・ Personal Vehicle Ownership: 2
(2002) types (owner, non owner)
・ Assigned traffic on the all OD
pairs
Modal Share ・ Simple Model ・ Estimated by disaggregated
・ “A priori” modal share logit model
・ Various modal share by
different transport policy
Traffic Assignment ・ Incremental assignment ・ Equilibrium assignment using
using daily capacity hourly capacity
・ 24 hours ・ AM peak hour, PM peak hour
・ Bus speed is fixed. and off-peak hour
・ Bus speed is affected by traffic
congestion.
Timing of Usage in the ・ Formulation and ・ Formulation and Evaluation of
Study Evaluation of Master F/S Projects
Plan
Application ・ JICA STRADA (CUBE) ・ CUBE/Voyager
Source: Study Team

3-1
 The Study on Urban Transport Master Plan and Feasibility Study in HCM Metropolitan Area (HOUTRANS)
Final Report
VOLUME 4: Transport Database
3.2 2nd Stage Model in CUBE/Voyager
1) General Description
The fundamental structure of the 2nd stage model in CUBE/Voyager for the years 2002
and 2020 are shown in Figure 3.2.1.
Figure 3.2.1 HOUTRANS Base Year 2nd Stage Model in CUBE/Voyager Software

Source: Study Team

3-2
 The Study on Urban Transport Master Plan and Feasibility Study in HCM Metropolitan Area (HOUTRANS)
Final Report
VOLUME 4: Transport Database
Figure 3.2.1 HOUTRANS Base Year 2nd Stage Model in CUBE/Voyager Software
(continued)

Source: Study Team

2) Development of Base Year Trip Matrices and Network

An integral part of the 2nd Stage model was the development of the base year trip matrices
by mode and the highway network. For the CUBE/Voyager model these were transferred
directly from the JICA STRADA model in formats acceptable to the relevant programs in
CUBE/Voyager software. The base year study area network required for modelling was
built from the highway network used by the HOUTRANS STRADA model. The STRADA
model network was converted to an ASCII format and then read in by the relevant
CUBE/Voyager program. This process is illustrated in Annex 2.1.

The trip matrices were directly built from the final output files of the HIS (HIS) and the
cordon survey result files, to obtain trips made by study area residents and non-residents
(those travelling from outside) respectively. This process is detailed in Annex 2.2 to this
Report.

3) Trip Generation and Attraction Models

Trip Generation and Attraction models developed were of multiple linear forms as
illustrated by the equations below, the calibrated parameters and associated statistics are

3-3
 The Study on Urban Transport Master Plan and Feasibility Study in HCM Metropolitan Area (HOUTRANS)
Final Report
VOLUME 4: Transport Database
given in Table 3.2.1. The overall study area wide trips are then controlled to a total of
40.5m trips, implying a trip rate of 3.0 for the 2020 population of 13.5 million.

Generations: Gi = a*POP + b* WRK + c*SCL + C

Attractions: Ai = x*POP + y* WRK + z*SCL + D

Where: POP : Population in a Zone

WRK : Number of workers at work place in a Zone
SCL : Number of students at school in a Zone

a,b,c,x,y,z : Parameters
C&D : Constants

Table 3.2.1 Calibrated Parameters for Trip Generation and Attraction Models

Generation Attraction
Variable
Coefficients t-value Coefficients t-value

POP=Population a=1.404 14.5 x=1.429 14.5

WRK=No. of workers at workplace b=2.627 12.8 y=2.524 12.1

SCL=No. of students at school c=2.284 7.1 z=2.333 7.1

Constants C & D C=-1054.0 -0.4 D=-869.5 -0.3

Correlation Coefficient 0.931 0.928

Source: Study Team

Calculations for trip Generations and Attractions were performed using CUBE/Voyager
program “GENERATIONS,” the process is detailed in Annex 2.3.

4) Trip Distribution Models

The Trip distribution models developed for the HOUTRANS study were of two types, and
applied in two stages. In the first stage Intrazonal trips were estimated and then the
Interzonal trips are computed using conventional gravity type model. The detailed
implementation of these models is illustrated in Annex 2.4.

1. Intrazonal Trip Model

Intrazonal trips in a zone were estimated from the synthesized trip Generations and
Attractions by using the observed proportions of Intrazonal trips in the base year. For the
future years the proportion of Intrazonal trips were “adjusted”. The form of the model is:

Tii = I ii (Gi + Ai ) / 2
Where: Tii : No. of Intrazonal trips in zone i
I ii : Proportion of Intrazonal trip in zone i derived from HIS surveys;
i.e. I ii = {(Observed Intrazonal Trips in Zone) /[0.5*(Gensi+Attsi)]}

For some zones outside the HCMC urban area, the proportion of present Intrazonal trips

3-4
 The Study on Urban Transport Master Plan and Feasibility Study in HCM Metropolitan Area (HOUTRANS)
Final Report
VOLUME 4: Transport Database
were observed to be high, due to the short trip lengths in HCMC compared with other
similar Asian cities. Therefore, the future Interzonal proportions were derived taking into
account the size of the zones and the network connectivity – as these factors influenced
the self-containment of an area. The base and forecast year proportions are given in Table
3.2.2.
Table 3.2.2 Interzonal Trip Ratios by Area Types – Observed and Future

Area Intrazonal Trip Ratio

Area Description Zones
(ha) Observed Adjusted
As in Base
Inner Core 91 46 Average=0.19
Year Prop.
Inner Fringe 45 206 0.32 0.25
HCMC Emerging
56 989 0.50 0.45
Peripheral
Suburban 17 1,077 0.45 0.40
Rural 7 12,745 0.70 0.60
Satellite Urban 17 1,826 0.64 0.55
Other
Suburban 4 3,473 0.70 0.60
Provinces
Rural 28 8,828 0.83 0.70
Study Area Total 265 1,768 0.41
Source: Study Team

2. Interzonal Trip Distribution Model

After estimating the Intrazonal trips, the Interzonal trips were estimated by using a gravity
model. The development model took the following form, and the calibrated parameters are
given in Table 3.2.3.

Giα * A βj
Tij = κ (i ≠ j )
d ijγ
Where: Tij : No. of Interzonal trips between zone i and j
d ij : Interzonal Distance between zone i and j
k, α, β , γ : Calibrated Parameters

Table 3.2.3 Calibrated Parameters of Interzonal Trip Distribution Model

Parameters Coefficients t-statistics

α 0.5840 43.0
β 0.5980 44.0
γ 0.7280 76.9
κ 0.0027 31.2
Correlation Coefficient 0.51
Source: Study Team

Figure 3.2.2 shows the results of calibrated distribution model. Regarding the trip length
distribution, chi-squire value resulted to 0.0, which implies that the fitness of observed and
estimated value is good fit and is significant.

3-5
 The Study on Urban Transport Master Plan and Feasibility Study in HCM Metropolitan Area (HOUTRANS)
Final Report
VOLUME 4: Transport Database
Figure 3.2.2 Comparison of Observed and Synthesized Trip Length Distribution

2,500

2,000

No. of Trips (000) 1,500 Observed

1,000 Estimated

500

0
0 5 10 15 20 25 30 35
Travel Distance (km)
Source: HOUTRANS
Source: Study Team

5) Modal Split Models

The number of trips distributed comprised: walk trips and those made by five mechanized
modes of transport: bicycle, motorcycle, car, truck and bus. The modal split model
developed was a 2-stage process.

In the first step walk trips were estimated by applying the proportion of walk trips to total
trips observed between each zone pair in the base year. These trips were then removed
from the total trip matrix to yield total mechanized mode trips.
In the second step, the remaining five modes were estimated using a Logit function based
on the generalized cost of travel by each mode between all zones pairs. The Intrazonal
trips were removed before the modal split model as these trips did not serve any further
purpose in the demand forecasting process.

The form of the Logit model calibrated to estimate trips by each of the five mechanized
mode was:

1
Pijm = αC m + β
1 + e ij
s.t. ∑ Pijm = 1
m∈M

d ij * VC m + PC m
C = t * VT +
m
ij
m
ij
m

Occ m
Where: Pijm : Probability of choosing mode m between zone i and j
C ijm : Generalized cost of mode m between zone i and j
t ijm : Travel time by mode m between zone i and j
VoT m : Value of Time of user of mode m
VoC m : Vehicle operating Cost of mode m (for bus, fare)
PC m : Parking cost of mode m
Occ m : Average Vehicle Occupancy of mode m
α, β : Calibrated Parameters

3-6
 The Study on Urban Transport Master Plan and Feasibility Study in HCM Metropolitan Area (HOUTRANS)
Final Report
VOLUME 4: Transport Database
The model was calibrated. The resulting model parameters are given in Table 3.2.4 and
Table 3.2.5 below.
Table 3.2.4 Modal Split Model Generalized Cost Parameters by Mode

Value of Time Vehicle Average

VoT ($/hour) Parking
Operating Occupancy
Mode Fare ($) Cost
Cost (Person/
2002 2020 ($/Trip)
VoC($/km) Vehicle)
Bicycle 0.464 n/a 0.0034 - 0.039 1.14
Motorcycle 0.823 2.52 0.0268 - 0.083 1.51
Car 1.723 5.26 0.0705 - 0.314 1.96
Truck 1.025 n/a 0.1535 - 0.350 1.98
<=4km:0.067,
Bus 0.525 2.52 - 4km>: - -
0.067+0.016/km
Source: Study Team
n/a=for 2020 it was assumed that no trios will be made by these modes of transport
Other GC parameters were assumed to remain same for 2020, i.e. no increase in real terms.

Table 3.2.5 Calibrated Modal Split Model Parameters

Mode Parameters
Multiple Correlation
α β
Bicycle 0.598 1.814 0.87
Motorcycle 0.829 -1.907 0.97
Car 0.139 2.950 0.62
Truck 0.133 3.963 0.97
Bus 0.513 3.052 0.64
Source: Study Team

The HOUTRANS mode split model calibration results had been reported in earlier reports
and Technical Reports on modeling, and therefore are not repeated here. The model
implementation in CUBE/Voyager form has been detailed in Annex 2.5 to this Report.

6) Traffic Assignment Model

The modal split model yields the number of person trips made by each mode of travel.
Prior to the assignment, these person trips are converted to their equivalent Passenger
Car Units (PCU) trips. The factors used for this conversion are given in Table 3.2.6 for both
base and forecast years.

The assignment procedure adopted was based on capacity restraint process. The
24-traffic volume was assigned to the network using “Incremental” loading process over
10 iterations, assigning 10% of the traffic in each iteration. A link time functioning as path
building parameter was found to give the best validation results. The link times were
adjusted to take into account the reduction in speed based on link capacity. The speed
flow relationship applied was the same for all links in the network. This is illustrated below
in Figure 2.7. The speed on all centroid connectors was assumed to stay constant at
5kph.

The complete Assignment process is illustrated in Annex 2.6 to this Report.

3-7
 The Study on Urban Transport Master Plan and Feasibility Study in HCM Metropolitan Area (HOUTRANS)
Final Report
VOLUME 4: Transport Database
Table 3.2.6 Base and Forecast Year Vehicle Occupancy and PCU Factors

Vehicle Occupancy PCU Factor

Mode
2002 2020 2002 2020
Bicycle 1.14 - 0.20 -
Motorcycle 1.51 1.30 0.20 0.30
Car 1.96 1.96 1.00 1.00
Truck 1.98 - 2.50 2.50
Bus 8.18 50.0 2.50 2.50

Source: Study Team

Figure 3.2.3 Speed Flow Relationship Used for All Links in Assignment Model

Free Flow Speed

Speed

Min Speed = }
0.1*FF Speed}
0 Traffic Volume Q=Capacity

Source: Study Team

3-8
 The Study on Urban Transport Master Plan and Feasibility Study in HCM Metropolitan Area (HOUTRANS)
Final Report
VOLUME 4: Transport Database
3.3 3rd Stage Model
1) General Description
This model is a disaggregated version of the 2nd Stage model. However, the general form
of the model is still the conventional four-step. The model was developed using JICA
STRADA, and later enhanced, when possible, and converted to the CUBE/Voyager
software. The main four steps of the model are:

1. Trip Generation and Attraction Models – to estimate the number of trips generated by
and attracted to each zone. In this 3rd Stage model the trip-end was estimated by five
trip purposes and seven household categories.

2. Trip Distribution Models – to estimate the number of trips traveling within and between
zones. Trips distribution was carried out separately by trip purpose and household
category.

3. Modal Split Models – to estimate the number of trips by each mode of travel. Again the
modal split models were applied by trip purpose and household category for walk
mode and further four mechanized modes (bicycle, motorcycle, car and public
transport(by bus or rail).
4. External Trip Model – to estimate the trips made by the non-residents of the study
area, within the study area and across study area cordon.

5. Peak Period Models – to estimate trips by highway and public transport traveling
during the morning, evening and off-peak time periods; and
6. Traffic Assignment Models - to estimate the number of trips traveling on roads for
private (bicycle, motorcycle, and car) and public transport mode in each of the three
time periods.

Its should be noted that in the case of the 3rd Stage model no base year model
development and calibration is described here, as the conversion to CUBE/Voyager was
only carried out to run the models for the forecast year test scenarios. However, the peak
period models developed at the later stage using CUBE/Voyager are described in this
document.

The full 4-stage process is shown next in Figure 3.1.1. It illustrates how the model
becomes more complex. The first page, in Figure 3.1.2 depicts the general form of the
model. Figure 3.1b depicts the model in the CUBE/Voyager platform. Each of the main
component’s model: trip-end, trip distribution, modal split, peak period, external trip model,
Highway & PT assignments are represented by a separate subgroup. The figure also
illustrates the linkages of data output by one model and fed into the next model.
The remaining sections of this chapter describes in detail the submodals of the 3rd Stage
model as implemented in CUBE/Voyager platform.

3-9
 The Study on Urban Transport Master Plan and Feasibility Study in HCM Metropolitan Area (HOUTRANS)
Final Report
VOLUME 4: Transport Database
Figure 3.3.1 HOUTRANS 3rd Stage Model – Overview of Model Configuration

Socio- Economic
Framework

Vehicle
Trip Control Vehicle Ownership
Ownership
Total Model by Household
Model

Study Area
Trip Control
Totals

Generation
Highway Network Vehicle Ownership
Attraction
by Zone
Models

Base Assignment Trip Generation &

Attraction by Purpose &
Household Categories

Trip Distribution Observe External

Travel Time Travel Distance Trip Matrices
Model

Travel Speed & Trip Matrices by External Trips

Cost Model Purpose & HH Model
Type

Modal Split
Travel Speed & Cost Model

24HR External
24HR Trip Matrices Peak Model Trip Matrices by
by Mode & Purpose Mode

Initial Highway Assignment PCU

Peak Total Matrices
Assign Private PCU Matrix
by Mode

PT Next Work Model with

Adjusted Highway Speeds for
Buses Legend

Inputs
PT Person Trips
PT Assignment Model

Final Highway
Trip
Assignment Model
Maxtrix

Final Highway Model

PT Volume by Mode and Line PT Pre- Loads
Assignment

Evaluation

Source: Study Team

3-10
 The Study on Urban Transport Master Plan and Feasibility Study in HCM Metropolitan Area (HOUTRANS)
Final Report
VOLUME 4: Transport Database
Figure 3.3.2 HOUTRANS 3rd Stage Model – Overview in CUBE/Voyager Platform

Source: Study Team

2) Trip Generation and Attraction Models

The 3rd Stage trip-end model is a two-stage process, in the first stage the zonal trip
generation and attractions were estimated for each of the five trip purposes according to
the household vehicle ownership category. The zonal trip-ends were then controlled to the
overall trips estimated exogenously at the study area level by the five trip purposes and
seven household categories.

The following five trip purposes were modeled in such a way that the trip purpose was at
the destination end of the trip, irrespective of the purpose at the trip origin. The estimated
trip-ends are therefore trip origins and destinations, and are not the conventional trip
generations and attractions (even though they are referred as such) which are usually
represented at home and non-home end of the trips.

3-11
 The Study on Urban Transport Master Plan and Feasibility Study in HCM Metropolitan Area (HOUTRANS)
Final Report
VOLUME 4: Transport Database
The trip purposes and household categories are listed below:

Modeled Trip Purposes

1. To Work
2. To School

3. Private (e.g. Shopping, Leisure, or Personal purposes etc)

4. Business (on Official Business)

5. To Home (i.e. from any where returning home)

Modeled Household Categories

1. Car Owning Household – Proportion of the HH Population with Vehicle Available for
travel

2. Car Owning Household – Proportion of the HH Population with No Vehicle Available for
travel

3. 2+ Motorcycle (M/C) Owning Household – Proportion of the HH Population with

Vehicle Available for travel
4. 2+ Motorcycle (M/C) Owning Household – Proportion of the HH Population with No
Vehicle Available for travel
5. 1 Motorcycle (M/C) Owning Household – Proportion of the HH Population with Vehicle
Available for travel

6. 1 Motorcycle (M/C) Owning Household – Proportion of the HH Population with No

Vehicle Available for travel
7. Non Vehicle Owning Households

The estimated trip-ends therefore require no further processing to convert them to

origin/destination form. The calibrated zonal trip-end model was determined to be a
multiple linear type and may be described as:

= ∑ a snk ⋅ X sin + c nk = ∑ a snk ⋅ X sjn + c nk

nk nk
Gi and Aj
s s

Where: Gink , A nk
j : trip generation from zone i or trip attraction to zone j by
person of household category n with trip purpose k.

X sin : Number of persons of household group n in zones i estimated

exogenously from the planning data, and a snk , c nk : calibrated parameters

The calibrated parameter values for the trip generations and attractions are given in
Tables 3.3.1 and Table 3.3.2, respectively.

3-12
 The Study on Urban Transport Master Plan and Feasibility Study in HCM Metropolitan Area (HOUTRANS)
Final Report
VOLUME 4: Transport Database
Table 3.3.1 Trip Generation Model Parameters

Household Category
Trip
Variables 2+ M/C 1 M/C Non-
Purpose Car Owning
Owning Owning motorized
Resident primary workers – – 1.292 –
Resident secondary workers – 1.306 1.150 –
To Work Resident tertiary workers – 1.179 0.875 –
Total Resident workers 1.392 – – 1.054
Constant 33.1 31.9 28.6 –
Resident primary students – 1.331 1.042 –
Resident secondary students – 1.243 1.102 –
To School Resident tertiary students – 0.927 1.388 –
Total Resident students 1.195 – – –
Constant – – – 1.076
Population 0.989 0.726 0.667 0.668
Private Total Workers=Jobs 1.125 0.331 0.237 –
Constant – – – 106.5
Secondary Workers=Jobs – 0.072 0.035 –
Business Tertiary Workers=Jobs 0.787 0.154 0.104 0.053
Constant 55.5 32.2 178.6 22.9
Population 0.077 0.269 0.348 0.673
Tertiary Workers=Jobs – 2.667 – –
To Home
Total Worker=Jobs 3.152 – 2.256 1.456
Constant 248.1 – 1298.2 51.9
Source: Study Team

Table 3.3.2 Trip Attraction Model Parameters

Household Group
Trip
Variable 2+ M/C 1 M/C Non-
Purpose Car Owning
Owning Owning motorized
Primary Workers=Jobs 0.174 – 1.228 1.453
Secondary Workers=Jobs 2.682 1.356 1.202 0.995
To Work
Tertiary Workers=Jobs 1.057 1.276 0. 951 0.791
Constant 53.7 39.5 – 6.3
Primary School Places – 0.997 1.085 –
Secondary School Places – 1.367 1.176 –
To School Tertiary School Places – 1.286 1.000 –
Total School Places 1.115 – – 1.108
Constant 10.0 – – –
Population 0.989 0.340 0.308 0.510
Private Tertiary Workers=Jobs 1.125 1.846 2.166 0.580
Constant 159.1 – 708.0 185.5
Population – – – 0.008
Secondary Workers=Jobs – 0.097 0.051 –
Business Tertiary Workers=Jobs – 0.153 0.097 –
Total Workers=Jobs 0.492 – – –
Constant 65.9 – 171.4 21.7
Population 2.040 1.604 1.403 1.225
To Home Total Resident Workers 0.075 – – 0.238
Constant 2.1 – 604.9 –
Source: Study Team

3-13
 The Study on Urban Transport Master Plan and Feasibility Study in HCM Metropolitan Area (HOUTRANS)
Final Report
VOLUME 4: Transport Database
The above produced zonal trip-ends for the study area internal zones (1-265). These
zonal trip-ends were then controlled to the overall study area trips estimated at the study
area level using the calibrated trip rates for each household category. The trip rates are
given below in Table 3.3.3.

Total attractions were then balanced to the total controlled generations at the study area
level.

After the application of trip control totals by trip purpose and by household categories, the
total study area level trips were then controlled to an average trip rate of 3.0 trips per
person. Therefore, for the projected 2020 population of 13.5 million, the total trips were
controlled to 40.5 million trips per day. The trip-end models as implemented in
CUBE/Voyager are further illustrated in Annex 3.1.
Table 3.3.3 Trip Rate for Trip Generation Control Totals at Study Area Level

Vehicle
Household Group To Work To School Private Business To Home
Av / NVA
Vehicle Av 0.879 0.172 1.466 0.406 2.125
Car Owning HH Pop
Vehicle N/A 0.190 0.485 0.792 0.079 1.341

Vehicle Av 0.757 0.214 0.863 0.069 1.615

2+ M/C Owning HH Pop
Vehicle N/A 0.180 0.378 0.718 0.020 1.207

Vehicle Av 0.639 0.208 0.926 0.067 1.562

1 M/C Owning HH Pop
Vehicle N/A 0.211 0.364 0.737 0.017 1.235

Non-motorized HH Pop N/A 0.420 0.258 0.716 0.023 1.289

Source: Study Team

Vehicle Available refers to household population who has available vehicle; NVA to household population
which does not have an available vehicle for travel, even though they are vehicle owning HH population.

3) Trip Distribution Models

Similar to the 2nd Stage, a two step trip distribution process was applied. In the first step
Intrazonal trips were estimated based on proportion of trip generations and attractions. In
the second step Interzonal trips were computed using conventional gravity type model.
The 2nd step of the trip distribution model developed for the 3rd Stage modeling involved
separately distributing trips for each trip purpose and by each household category. In total
there were 35 trip distribution models at the 2nd Step, compared to a single model
representing all trips, as used in the 2nd stage model.
1. Intrazonal Trip Estimation Model

Intrazonal trips in a zone are a function of total trips generated and attracted to a zone.
These were estimated from the synthesized trip Generations and Attractions by using the
observed proportions of Intrazonal trips for the inner area of the HCMC study area.
However, for out areas, where the trip pattern is likely to change in the future, the base
year proportions were adjusted to take into account the lessening of the self-containment
of the area and the expected increase in trip lengths with the increase in incomes and the
availability of vehicles. For future years the proportion of Intrazonal trips were “adjusted”,
and are tabulated below. The form of the model is:

3-14
 The Study on Urban Transport Master Plan and Feasibility Study in HCM Metropolitan Area (HOUTRANS)
Final Report
VOLUME 4: Transport Database

Tii = I ii * min(GiorAj )
Where: Tii : No. of Intrazonal trips in zone i
I ii : Proportion of Intrazonal trip in zone i derived from HIS
surveys;
i.e. I ii = {(Observed Intrazonal Trips in Zone) /[0.5*(Gensi+Attsi)]}

Gi : Future Trip Generations of Zone i.

Aj : Future Trip Attractions of Zone j.

For some zones outside the HCMC urban area, the proportion of present Intrazonal trips
were observed to be high, due to the short trip lengths in HCMC compared with other
similar Asian cities. Therefore, the future Interzonal proportions were derived taking into
account the size of the zones and the network connectivity – as these factors influence the
self-containment of an area. The base and forecast year proportions are given in Table
3.3.4, and Table 3.3.5 defines the areas.

Table 3.3.4 Interzonal Trip Ratios by Area Types – Observed and Future

Area Intrazonal Trip Ratio

Area Description Zones
(ha) Observed Applied
Average As in Base
1. Inner Core 91 46
=0.19 Year Prop.
2. Inner Fringe 45 206 0.32 0.25
HCMC 3. Emerging Peripheral 56 989 0.50 0.45
4. Suburban 17 1,077 0.45 0.40
5. Rural 7 12,745 0.70 0.60
6. Satellite Urban 17 1,826 0.64 0.55
Other
7. Suburban 4 3,473 0.70 0.60
Provinces
8. Other Rural 28 8,828 0.83 0.70
Study Area Total 265 1,768 0.41 0.32
Source: Study Team

Table 3.3.5 Area Definition for the Application of Intrazonal Trip Proportions

Area Area Name Districts Zones Zones Zones

1 HCMC Inner Core 1, 3-6, 10-11, 16 1-59 73-97 130-136
2 HCMC Inner Fringe 8, 13-15 60-72 98-129
3 HCMC Peripheral 2, 7, 9, 12, 17, 19 137-163 171-190 201-209
4 HCMC Suburban 18, 20 164-170 191-200
5 HCMC Rural 21-22 210-216
6 Satellite Urban 23, 26, 29 217-218 240-243 248-258
7 Satellite Suburban 24-25 244-247
8 Other Rural 27-28, 30-36 219-239 259-265
Source: Study Team

3-15
 The Study on Urban Transport Master Plan and Feasibility Study in HCM Metropolitan Area (HOUTRANS)
Final Report
VOLUME 4: Transport Database
2. Interzonal Trip Distribution Model

After estimating the Intrazonal trips, the Interzonal trips are estimated by using a gravity
model. The model developed took the following form, and the calibrated parameters are
given in Table 3.3.6.

Giα * A βj
Tij = κ (i ≠ j )
d ijγ
Where: Tij : No. of Interzonal trips between zone i and j
d ij : Interzonal Distance between zone i and j
k, α, β , γ : Calibrated Parameters.

After the estimation of synthesized trips, the Fratar balancing process was applied to
balance the distributed trip totals to zonal generations and attractions, with final balance to
generations, such that the difference between trip attractions was minimized. The
complete process is depicted in Annex 3.2.

Table 3.3.6 on the next page shows the similarities in parameters for both private and
business trips. Due to the small business sample size, it was not possible to calibrate
separate parameter values. Therefore the combined model was calibrated. However in
order to maintain the highest level of disaggregation the models were distributed using the
same parameters for private and business trips, with the exception of Non-vehicle-owning
household trips.

4) Modal Split Models

The third step of the travel demand forecast procedure ‘modal split’ model estimates by
which mode of travel, the origin-destination volumes estimated by the trip distribution
model will use. In some models, modal split step was carried out before the trip
distribution stage. In this case, modal share was determined reflects only the function of
modal attributes at the origin or the destination ends, and not the level of modal services
between origin and destination zones. Therefore, modal split if performed after the trip
distribution would show further advantages, and this approach was adopted for the
HOUTRANS model.
Similar to the 2nd Stage, a two step modal split process was applied. In the first step walk
trips, which have limited range of travel time & distance, and have no out-of-pocket cost of
travel (other than the time value) were estimated. These trips were then subtracted from
the total distributed trips. The remainder trips were then subjected to full mode choice
model based on cost and other attributes of travel by each mode. This process was
applied by each trip purpose and for each household category to estimate trips by the
following four main modes of travel.

3-16
 The Study on Urban Transport Master Plan and Feasibility Study in HCM Metropolitan Area (HOUTRANS)
Final Report
VOLUME 4: Transport Database
Table 3.3.6 HOUTRANS 3rd Stage Interzonal Trip Distribution Parameters

Household Category Trip Purpose K Alpha(α) Beta(β) Gama(γ)

Work 1.390 0.332 0.309 0.0179
School 2.780 0.413 0.190 0.0298
1. Car Owning HH,
Private 0.968 0.414 0.320 0.0412
Car Available
Business 0.968 0.414 0.320 0.0412
Home 1.390 0.259 0.343 0.0173
Work 3.020 0.329 0.316 0.0261
School 0.859 0.497 0.324 0.0375
2. Car Owning HH,
Private 3.220 0.310 0.287 0.0301
Car Not Available
Business 3.220 0.310 0.287 0.0301
Home 3.150 0.268 0.263 0.0206
Work 0.732 0.326 0.266 0.2900
School 3.590 0.249 0.164 0.1150
3. 2+ M/C Owning HH,
Private 1.310 0.321 0.224 0.1790
M/C Available
Business 1.310 0.321 0.224 0.1790
Home 0.247 0.334 0.368 0.3400
Work 0.634 0.283 0.369 0.0289
School 1.290 0.338 0.197 0.2440
4. 2+ M/C Owning HH,
Private 0.793 0.303 0.327 0.1320
M/C Not Available
Business 0.793 0.303 0.327 0.1320
Home 0.362 0.315 0.357 0.2710
Work 0.434 0.407 0.297 0.1440
School 1.760 0.472 0.162 0.0323
5. 1 M/C Owning HH,
Private 0.290 0.474 0.303 0.0210
M/C Available
Business 0.290 0.474 0.303 0.0210
Home 0.112 0.359 0.499 0.2120
Work 0.382 0.377 0.357 0.0481
School 0.291 0.396 0.357 0.2200
6. 1 M/C Owning HH,
Private 0.127 0.467 0.406 0.1740
M/C Not Available
Business 0.127 0.467 0.406 0.1740
Home 0.025 0.501 0.505 0.3010
Work 1.210 0.298 0.339 0.0812
School 1.620 0.272 0.333 0.0435
7. Non Vehicle Owning HH Private 0.882 0.359 0.332 0.0220
Business 0.794 0.538 0.447 0.1010
Home 0.449 0.380 0.349 0.1450
Source: Study Team

3-17
 The Study on Urban Transport Master Plan and Feasibility Study in HCM Metropolitan Area (HOUTRANS)
Final Report
VOLUME 4: Transport Database
1. Trips by Bicycle
2. Trips by Motorcycle
3. Trips by Car, and
4. Trips by Public Transport (PT or some time referred to as Bus trips).
The choice of bus vs a metro is estimated at the public transport assignment stage. The
complete modal split model process is shown in Figure 3.3.3 below.
Figure 3.3.3 Modal Split Model Process

Source: Study Team

Walk Modal Split Model The calibrated model took the form of an exponential based on
distance traveled. Two models were developed, as choice to walk was quite dependent on
vehicle availability. The HIS analyses showed that 98% of all walk trips were under 5km.
Therefore, for the forecast year the maximum walk distance was set at 5km. The general
form of the model was:

1
=
lk
Pij α ⋅d ij + β
1+ e
Where: Pijlk : share of walk trips by person with personal vehicle available l,
for trip purpose k between zones of i and j,

d ij : travel distance between zones of i and j. For Intrazonal trips the

radius of the zone was taken as the distance traveled, and

α and β are the calibrated parameters depending upon the vehicle

availability.

3-18
 The Study on Urban Transport Master Plan and Feasibility Study in HCM Metropolitan Area (HOUTRANS)
Final Report
VOLUME 4: Transport Database
Table below list the calibrated values by trip purpose and vehicle available group.
Table 3.3.7 Parameters for Walk Modal Split Model

Household Group Parameters To Work To School Private Business To Home

Vehicle Owning & Coefficient ( α ) 0.364 1.335 0.512 1.151 0.364

Vehicle Available Constant ( β) 2.076 1.324 1.294 1.467 1.781

Vehicle Not Coefficient ( α ) 0.289 0.254 0.239 0.170 0.344

Available Constant ( β) 0.201 0.412 0.064 0.787 0.240
Source: Study Team

Main Modal Split Model: This model divides the remaining non-walk trips into four modes
of travel as described above. An aggregate logit model has been applied and calibrated.
In this model the random terms ( ε ) in the utility function, U = V + ε , are identically and
independently distributed with Gumbel double exponential probability distribution function.
This model is expressed below:

mk
exp(V ij )
=
mk
P
∑ exp(V ij )
ij lk

l
Where:
Pijmk is the share of trips travel by mode m, with purpose k, between zones of i and j.
Vijmk , Vijlk are the utility of trips travel by mode m or l with purpose k between
zones of i and j, and the utility function can be expressed as follows:

= β1lk ⋅ Z 1ij + β 2lk ⋅ Z 2ij + L + β nlk ⋅ Z nij

lk lk lk lk
V ij
Where: Z nij
lk
: variable n for mode l and trip purpose k., and
β nlk : parameter for variable n.

The calibrated parameters for five trip purposes and four modes are given in Table 3.3.8.

3-19
 The Study on Urban Transport Master Plan and Feasibility Study in HCM Metropolitan Area (HOUTRANS)
Final Report
VOLUME 4: Transport Database
Table 3.3.8 Modal Split Model Parameters

Mode Specific Parameter Values by Trip Purpose

Mode Variable
Work School Private Business Home
Bicycle VOC (0.0069) (0.0151) (0.0116) - (0.0108)
Car Owning - HH - - - - -
2+ M/C – HH - - - - -
1 M/C – HH 1.7450 2.0072 2.4524 - 2.2113
Bicycle

Non Motorized HH 2.3793 2.1256 1.2833 - 1.6730

Vehicle Available Y/N (0.9700) - (2.7932) - (1.0446)
Time Parameter - - - (4.0144) -
Distance Parameter (0.1500) (0.1500) (0.1500) (0.1500) (0.1500)
Mode Specific Constant - B/C 0.7000 0.7000 0.7000 0.7000 0.7000
Motorcycle VOC (0.0010) (0.0025) (0.0018) - (0.0017)
Car Owning - HH - - - - -
2+ M/C - HH 1.1726 0.9392 2.2827 - 1.3537
Motorcycle

1 M/C – HH 1.5450 1.8338 2.7616 - 2.1930

Non Motorized HH - - - - -
Vehicle Available Y/N 2.9246 3.5250 - - 2.4340
Time Parameter - - - (3.7722) -
Mode Specific Constant - M/C (0.2600) (0.2600) (0.2600) (0.2600) (0.2600)
Car VOC (0.0003) (0.0009) (0.0006) - (0.0006)
Car Owning - HH 2.2974 - - - 2.2503
2+ M/C - HH - - - - -
1 M/C – HH - - - - -
Car

Non Motorized HH - - - - -
Vehicle Available Y/N - - - - -
Time Parameter - - - (6.4922) -
Mode Specific Constant - Car - - - - -
Bus_Fare (0.0010) (0.0036) (0.0023) (0.0002) (0.0023)
Car Owning - HH - - - - -
Public Transport

2+ M/C - HH (1.2991) - - - -
1 M/C - HH (0.9346) - - - -
Non Motorized HH - - - - -
Vehicle Available Y/N - - (1.8453) - (0.4152)
Time Parameter - - - (9.6750) -
Mode Specific Constant - PT 0.3250 0.3250 0.3250 0.3250 0.3250
Source: Study Team, VoC=Vehicle Operating Cost

Using the above parameter values and associated attribute values of each mode of travel
the probability of travel by each mode may be calculated as:
mode M mode M bicycle m/c car bus
Pij = exp(Uij ) / [exp(Uij ) + exp(Uij ) + exp(Uij ) + exp(Uij )]

Using these probabilities, the trips by each mode were estimated. The calculation of
mode specific vehicle operating cost (VoC), distance, time, parking charges, and other
parameter values used in the modal split are detailed in Annex 3.3. The annex also
details the likely costs values required to attain “a target modal shift” as used by the
HOUTRANS master plan study scenarios. A number of scenarios were tested, but only
the three key scenarios as discussed in the Master Plan Final report were documented.

3-20
 The Study on Urban Transport Master Plan and Feasibility Study in HCM Metropolitan Area (HOUTRANS)
Final Report
VOLUME 4: Transport Database
5) External Trip Model
The models described above estimate the trips made by the study area residents in the
internal study area (within and between Zones 1-265) only. In addition to these trips, the
trips made by the non-residents of the study area (within the study area (in zones 1-265)
and between the study area and 5-external zones (266-270) were estimated from the
observed cordon trip matrices using “Fratar” modeling technique. The growth factors
applied to the observed trips, by mode of travel. The following five modes were modeled:

• Bicycle
• Motorcycle
• Car
• Truck
• Public Transport (Buses)

The future year trips were based on the growth in the study area population, work places
and school places. The overall growth of these trips was capped to the total growth in
population of the study area. The same growth factors were used for each of the five
modes) of trips observed at the cordon. The CUBE/Voyager model detailed in Annex 3.4
gives the full process of estimation of the future external trip matrices.

6) Peak Period Models

The 2nd Stage model assignment was based on assigning all trips using daily capacity of
road network. It was found to be too coarse. The analysis of the observed trips by time of
day varies considerably by trip purpose and mode of travel.

It was observed that some 25% of all trips occur during morning peak of two hours
between 06:00 and 08:00, and within these two hours the trip purpose and mode choice
also differs considerably.
Similarly the evening peak was observed between 16:00 and 18:00 hours, and accounted
for close to 18% of all trips. The majority of the remaining trips occurred either between
the two peak periods or from 18:00 to 22:00. Thus close to 90% of all study area trips were
observed to be within the 16hour period from 06:00 to 22:00.
Figures 3.3.4 and 3.3.5 illustrate the distribution of trips and percentage of trips by time of
day and trip purpose, respectively. Similarly, Figures 3.3.6 and 3.3.7 illustrate the
distribution of trips and percentage of trips by time of day by mode of travel, respectively.

For detailed modelling of the 3rd Stage model, it was considered more appropriate and
reasonable to model each time period separately. This was carried out by dividing the
daily trip matrices output by the modal split model into each time period by purpose and
mode.

It was observed that there were considerable differences in the proportion of trips
travelling during the peak or off peak hours in the direction of travel. This was analysed by
examining the proportion of trips to/from the CBD and other areas. In the final analyses
the study area was divided into 6 sectors as given in Table 3.3.9 below.

3-21
 The Study on Urban Transport Master Plan and Feasibility Study in HCM Metropolitan Area (HOUTRANS)
Final Report
VOLUME 4: Transport Database
Figure 3.3.4 Trip Distribution by Time of Day by Trip Purpose

Person Tips by Purpose by Time of day

2,500,000
Hom e
Business
2,000,000 Private
School
Work
No of Trips

1,500,000

1,000,000

500,000

0
1

3
-0

-0

-0

-0

-0

-1

-1

-1

-1

-1

-2

-2
00

02

04

06

08

10

12

14

16

18

20

22
Tim e of Day

Source: Study Team

Figure 3.3.5 % of Trip by Time of Day by Trip Purpose

% of Trips by Purpose and Time of Day

50.0%
Work
45.0% S c hool
40.0% P riva te
Busine ss
35.0% Home
% of Daily Trips

30.0% Tota l

25.0%
20.0%
15.0%
10.0%

5.0%
0.0%
01

03

05

07

09

11

13

15

17

19

21

23
-

-
00

02

04

06

08

10

12

14

16

18

20

22

Tim e of Day

Source: Study Team

3-22
 The Study on Urban Transport Master Plan and Feasibility Study in HCM Metropolitan Area (HOUTRANS)
Final Report
VOLUME 4: Transport Database
Figure 3.3.6 Trip Distribution by Time of Day by Mode of Travel

Person Tips by Mode of Travel by Time of day

3,000,000
PT
Car
2,500,000
M/C
Bicycle
2,000,000
No of Trips

1,500,000

1,000,000

500,000

0
1

3
-0

-0

-0

-0

-0

-1

-1

-1

-1

-1

-2

-2
00

02

04

06

08

10

12

14

16

18

20

22
Tim e of Day

Source: Study Team

Figure 3.3.7 % of Trip by Time of Day by Mode of Travel

% of Trips by Mode and Time of Day

20.0%
Byc yc le
18.0% Motorc yc le
Ca r
16.0%
PT
14.0% Tota l
% of Daily Trips

12.0%

10.0%

8.0%

6.0%

4.0%

2.0%

0.0%
01

03

05

07

09

11

13

15

17

19

21

23
-

-
00

02

04

06

08

10

12

14

16

18

20

22

Tim e of Day

Source: Study Team

3-23
 The Study on Urban Transport Master Plan and Feasibility Study in HCM Metropolitan Area (HOUTRANS)
Final Report
VOLUME 4: Transport Database
Table 3.3.9 Area Definition for the Application of Time Period Factors

Area Area Name Districts Zones

1 HCMC CBD 1, 3, 5, 10 1-24 33-47 73-84
2 HCMMC Inner Core 4, 6, 11, 16 25-32 48-59 85-97, 130-136
3 HCMC Inner Fringe 8, 13-15 60-72 98-129
HCMC Peripheral 2, 7, 9, 12, 17, 19 137-163 171-190 201-209
4
HCMC Suburban 18, 20 164-170 191-200
Satellite Urban 23, 26, 29 217-218 240-243 248-258
5
Satellite Suburban 24-25 244-247
HCMC Rural 21-22 210-216
6 Other Rural 27-28, 30-36 219-239 259-265
External Zones - 266-270
Source: Study Team

The final peak factors as applied to the forecast year trip matrices which are detailed in
Annex 3.5. It was not possible to maintain the level of disaggregation to 5 trip purposes
and 4 main modes of travel due to the small proportion of trips in some of the categories of
trips in a time period. Therefore, the final application was:
AM Peak

AM-Peak Hour= 1.95* AM-Peak period of 2 Hours from 06:00 to 08:00.

Area-to-area factors are to internal trip matrices output by mode split model by:
• 3 Modes: 1. 2-Wheel, 2. Car and 3-Public Transport Person Trips; and
• 4 Trip Purposes: 1-Work, 2-School, 3-(Private + Business), 4-Home

PM Peak
PM-Peak Hour= PM-Peak period of 2 Hours from 16:00 to 18:00.
Area-to-area factors are to internal trip matrices output by mode split model by:
• 3-Modes: 1. 2-Wheel, 2. Car and Public Transport Person Trips; and
• 2-Trip Purposes: 1-Non-Home (Work+School+Private+Business, 2-Home

Off Peak
Average Off-Peak Hour= 1/12 of the Periods from 08:00 to 16:00 and 18:00 to 22:00.
Area-to-area factors are to internal trip matrices output by mode split model by:
• 3-Modes: 1. 2-Wheel, 2. Car and Public Transport Person Trips; and
• 2-Trip Purposes: 1-Non-Home (Work+School+Private+Business), 2-Home

External Trips Peak Factors: For these trips trip factors were estimated using the cordon
and HIS database. Factors were estimated for all trip purposes that were combined, since
there was no trip purpose recorded on the cordon survey data. However, period factors
were developed for the following 4 modes of travel: 1-(Bicycle + Motorcycles) 2-Car,
3-Trucks and 4-PT.
The above time periods cover a period of 16-Hours, which represented about 90% of all
study area and cordon crossing trips. To obtain the daily (24-hour) traffic volumes in PCU’s
were analysed. It was concluded that these trips had similar travel patterns in the off peak
period, and amounted for about 7.5% of the 12-hour off-peak traffic volumes. Therefore,

3-24
 The Study on Urban Transport Master Plan and Feasibility Study in HCM Metropolitan Area (HOUTRANS)
Final Report
VOLUME 4: Transport Database
the daily trips for private PCU could be estimated from the above formulations. For the
public transport, it was observed that there was little bus or other para-transit operation
outside these hours. Therefore PT trips may be estimated using the above formulation
and the 24-hour trips are the same as for the 16-hour estimated by combining each of the
three period.

7) Traffic Assignment Models

In the 2nd Stage model a simplified traffic assignment process based on total (all-modes)
24-hour traffic volumes was used in conjunction with 24-hour road capacity values. This
process was useful when determining the overall demand for road space at an aggregate
level. However, it was not suitable for determining the mode specific requirements of
network/system capacities, both daily and by time of day.

For the 3rd stage modeling using CUBE/Voyager a peak period modeling approach was
adopted. This required the use of hourly network/system capacities. Road capacities were
obtained form the HOUTRANS database, where as likely future bus headways were
estimated depending upon the demand levels. The overall assignment process adopted
was for the three time periods: AM-Peak, PM-Peak and Off-Peak. For each time period
the process was comprised of the following steps, and a CUBE/Voyager application is
shown in three stages: 1. Initial assignment; 2. Pt Assignment & Pre-Loads and 3. Final
Highway Assignment.
Step 1: Combine post modal split trip matrices by mode and trip purpose for the study area
(Internal + External).

Step 2: Apply peak period factors by trip purpose and mode to study area and external
trips to obtain AM, PM and Average off-peak hourly trip matrices. See Annex 3.5.
Step 3: Apply vehicle occupancy and PCU factors to bicycle, motorcycle and car person
trips matrices output by the modal split model, and external trip models.

Step 4: Assign total PCU (2-Wheel+Car+truck) as a single user class to the highway
network, using time as route choice parameter.
Step 5: Adjust output congested highway speed for buses. Taken to be 70% of the
main-stream highway traffic speed on each link, and transfer these speeds to PT
assignment network.
Step 6: Assign total PT trips on the PT network with adjusted speeds. Bus-way and UMTR
speeds were set to 30kph, and minimum bus speeds to 5kph.
Step 7: Calculate bus passenger volumes on the network, convert to PCU and transfer to
highway network as pre-loads.

Step 8: Transfer any un-assigned PT trips from the PT un-assigned matrix to highway
matrix for assignment as para-transit PT trips in the highway assignment as
separate user class.

Step 9: Carry out final highway assignment as a multi-user class equilibrium assignment
for motorcycles, cars, Para-Transit PT volumes and bus volumes as pre-loads.

The above assignment process was carried out for three time periods. Traffic volumes

3-25
 The Study on Urban Transport Master Plan and Feasibility Study in HCM Metropolitan Area (HOUTRANS)
Final Report
VOLUME 4: Transport Database
from each time period assignments by each network link may then be analyzed to obtain
network performance indicators, or other economic, financial and environment analyses.

To obtain the daily volumes following factors were estimated:

Daily Private Traffic Volume = (1.95*AM-Peak + 2.0*PM-Peak+12.75* Off-Peak) Volumes.

Daily PT Volume = (1.95*AM-Peak + 2.0*PM-Peak+12.0* Off-Peak) Volumes

The highway assignment and the PT assignment may also be plotted for presentation
purposes using the CUBE/Viper program.

Speed/Flow Relationship in Equilibrium Assignment: The speed flow relationship used for
the 24-hour 2nd Stage model defined in section 2.7 and illustrated in Figure 2.7 was found
to be not suitable for the hourly assignments for each time period during the base year 3rd
Stage peak model calibration process. A new relationship as shown below in Figure 3.3.8
was found to give better peak period calibration.
Figure 3.3.8 Speed Flow Relationship for 3rd Stage Assignment Model

Speed Flow Relationship for Peak Period Assignment Models

50
Speed=FF Speed if Volume<=30% of Capacity
45
Free Flow Speed
40

35

30
Spe ed (kph)

25

20

15
Volume=Capacity
10
Minimum Speed= 10% of Free Flow Speed
5

0
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.1 1.2
Traffic Volume Expressed as Volume/Capacity Ratio

When testing a network scenario or any other policy assumptions the 3rd Stage model
outputs may differ from those produced using the 2nd Stage Model. Mainly because the
level of detail in the 3rd Stage model gives better estimates of likely future traffic volumes
by each mode than the aggregate volumes produced by the 2nd Stage model.

3-26
 The Study on Urban Transport Master Plan and Feasibility Study in HCM Metropolitan Area (HOUTRANS)
Final Report
VOLUME 4: Transport Database
3.4 Traffic Assignment Result of 3rd Stage Model
The traffic assignment on the Future Master Plan network was conducted to examine the
network performance by hourly assignment. As it was described, the 2nd stage model was
used for the M/P formulation and evaluation, however, the network performance is
calculated using the 3rd stage model (hourly assignment) here as reference.
Table 3.4.1, 3.4.2 and Figure 3.4.1, 3.4.2 show the result of traffic assignment using the 3rd
stage model. From this result, it was figured that there would be less traffic congestions
both during peak and off-peak hours in the public transport with a 50% case while the
traffic congestion is intense at the peak hours in the public transport with a 30% case.
These situations could be expressed only in the 3rd stage model.
1)
Table 3.4.1 Assessment of M/P Network: PT 30% Case

By Section By Area
Volume (000
Corridor/Area Capacity Volume (000 PCU/hr) V/C Ratio Capacity PCU-km/hr)
V/C Ratio
(000 (000
PCU/h) AM Off PM AM Off PM PCU-km/h) AM Off PM AM
Off peak
PM
peak peak peak peak peak peak peak peak peak peak peak
CBD Saigon 30.3 51.8 27.1 43.3 1.71 0.89 1.43 333 438 216 372 1.31 0.65 1.11
Cho Lon 44.4 52.5 30.5 46.4 1.18 0.69 1.04 322 339 161 298 1.05 0.50 0.92
I. NH1 Area 1 29.7 43.3 22.7 39.2 1.46 0.77 1.32 633 784 357 690 1.24 0.56 1.09
(East) Area 2 9.4 11.8 6.7 12.3 1.25 0.71 1.31 403 487 232 456 1.21 0.58 1.13
Area 3 9.8 11.6 6.0 11.7 1.18 0.61 1.19 424 416 169 405 0.98 0.40 0.96
Area 4 4.9 9.4 4.9 6.7 1.94 1.01 1.39 100 115 43 87 1.15 0.44 0.88
II. NH13 Area 1 4.1 5.2 2.4 5.1 1.26 0.58 1.24 147 171 73 162 1.16 0.49 1.10
Area 2 12.6 16.8 6.8 16.1 1.34 0.54 1.28 433 493 207 475 1.14 0.48 1.10
Area 3 12.1 8.2 3.9 8.2 0.68 0.32 0.68 219 155 71 161 0.71 0.32 0.73
III. NH22 Area 1 19.8 23.9 16.4 23.5 1.21 0.83 1.19 512 573 281 537 1.12 0.55 1.05
Area 2 15.9 22.7 12.2 23.0 1.42 0.76 1.44 506 588 306 593 1.16 0.61 1.17
Area 3 15.2 19.5 12.3 31.3 1.28 0.80 2.05 360 460 238 571 1.28 0.66 1.59
Area 4 14.6 11.0 5.1 11.1 0.75 0.35 0.76 411 370 186 442 0.90 0.45 1.08
VI. PR10 Area 1 19.7 17.7 10.2 16.4 0.90 0.52 0.83 164 152 65 130 0.93 0.40 0.79
Area 2 14.3 17.1 8.4 16.5 1.20 0.59 1.16 366 408 207 387 1.12 0.57 1.06
Area 3 13.3 10.0 3.9 10.7 0.75 0.29 0.81 412 397 179 408 0.96 0.43 0.99
Area 4 2.0 1.7 0.4 1.7 0.84 0.21 0.86 77 47 13 48 0.61 0.17 0.63
V. NH1 Area 1 13.0 11.8 7.0 11.2 0.91 0.54 0.86 50 37 19 34 0.74 0.37 0.68
(West) Area 2 7.8 5.5 2.6 5.8 0.71 0.33 0.74 132 138 77 131 1.04 0.58 0.99
Area 3 14.3 13.3 8.3 14.4 0.93 0.58 1.01 281 275 142 292 0.98 0.50 1.04
Area 4 5.1 5.4 5.1 5.3 1.05 0.99 1.03 583 351 186 407 0.60 0.32 0.70
VI. NH50 Area 1 16.9 15.5 7.6 15.2 0.91 0.45 0.90 142 154 76 139 1.09 0.54 0.98
Area 2 7.8 8.8 4.3 8.7 1.13 0.55 1.12 94 91 52 91 0.97 0.56 0.97
Area 3 4.9 6.3 4.6 7.1 1.29 0.94 1.45 527 368 122 441 0.70 0.23 0.84
Area 4 2.4 1.1 0.6 1.3 0.45 0.26 0.55 26 13 8 15 0.49 0.29 0.57
VII. PR15 Area 1 11.6 17.4 9.7 16.7 1.50 0.84 1.45 302 365 189 332 1.21 0.63 1.10
Area 2 6.9 9.6 6.0 10.8 1.40 0.87 1.56 45 58 35 61 1.27 0.77 1.34
Area 3 13.3 19.1 9.5 17.1 1.43 0.71 1.28 385 373 178 403 0.97 0.46 1.05
VIII. PR20 Area 1 1.9 2.3 1.1 2.1 1.19 0.57 1.08 125 143 79 127 1.15 0.64 1.01
(NH51) Area 2 8.1 9.2 5.5 9.0 1.14 0.68 1.11 146 138 71 134 0.95 0.49 0.92
Area 3 15.2 14.8 5.1 14.7 0.98 0.34 0.97 749 468 166 503 0.62 0.22 0.67
Area 4 5.9 4.6 2.3 4.5 0.78 0.38 0.76 465 98 43 103 0.21 0.09 0.22
Source: Study Team
1) Ave. occupancy: motorcycle - 1.3, car - 2.0, bus – 50
2) Toll fee on urban expressway: 15,000VND/ride; on interregional expressway: 1,000VND/km; fare on UMRT:
5,000VND/ride+500VND/km

3-27
 The Study on Urban Transport Master Plan and Feasibility Study in HCM Metropolitan Area (HOUTRANS)
Final Report
VOLUME 4: Transport Database
1)
Table 3.4.2 Assessment of M/P Network: PT 50% Case

By Section By Area
Capacity Volume (000
Corridor/Area Capacity Volume (000 PCU/hr) V/C Ratio
PCU-km/hr)
V/C Ratio
(000
(000
AM Off PM AM Off PM PCU-km/ AM Off PM AM Off PM
PCU/h)
peak peak peak peak peak peak h) peak peak peak peak peak peak
CBD Saigon 30.3 29.9 21.4 28.5 0.98 0.71 0.94 333 264 149 230 0.79 0.45 0.69
Cho Lon 44.4 38.1 22.8 31.1 0.86 0.51 0.70 322 214 112 171 0.67 0.35 0.53
I. NH1 Area 1 29.7 29.3 17.3 26.1 0.99 0.58 0.88 633 498 247 424 0.79 0.39 0.67
(East) Area 2 9.4 8.5 4.9 8.2 0.90 0.52 0.88 403 337 165 280 0.84 0.41 0.69
Area 3 9.8 8.6 4.2 6.9 0.88 0.43 0.71 424 281 129 195 0.66 0.30 0.46
Area 4 4.9 8.0 4.8 5.5 1.65 0.98 1.14 100 103 37 71 1.03 0.37 0.71
II. NH13 Area 1 4.1 3.6 2.3 2.9 0.88 0.57 0.70 147 105 48 87 0.71 0.33 0.59
Area 2 12.6 11.7 5.2 9.0 0.93 0.41 0.71 433 328 149 246 0.76 0.34 0.57
Area 3 12.1 6.7 3.3 4.9 0.55 0.27 0.40 219 110 56 88 0.50 0.26 0.40
III. NH22 Area 1 19.8 18.9 13.7 16.3 0.96 0.69 0.82 512 361 193 310 0.71 0.38 0.61
Area 2 15.9 15.1 8.8 12.5 0.95 0.55 0.78 506 384 227 322 0.76 0.45 0.64
Area 3 15.2 13.2 7.1 13.9 0.87 0.47 0.92 360 296 139 296 0.82 0.39 0.82
Area 4 14.6 7.6 3.4 6.5 0.52 0.23 0.44 411 239 115 256 0.58 0.28 0.62
VI. PR10 Area 1 19.7 12.0 8.1 10.4 0.61 0.41 0.53 164 82 45 66 0.50 0.27 0.40
Area 2 14.3 9.8 5.3 9.0 0.68 0.37 0.63 366 250 142 213 0.68 0.39 0.58
Area 3 13.3 4.6 2.2 5.0 0.35 0.17 0.38 412 229 105 199 0.56 0.26 0.48
Area 4 2.0 0.4 0.1 0.7 0.22 0.06 0.38 77 16 6 23 0.21 0.08 0.30
V. NH1 Area 1 13.0 9.2 5.3 7.2 0.71 0.41 0.56 50 23 12 19 0.45 0.25 0.38
(West) Area 2 7.8 3.5 1.7 3.0 0.45 0.22 0.38 132 95 57 81 0.71 0.43 0.61
Area 3 14.3 9.4 4.9 8.1 0.66 0.34 0.56 281 169 92 142 0.60 0.33 0.51
Area 4 5.1 5.3 4.5 5.2 1.03 0.87 1.01 583 244 128 204 0.42 0.22 0.35
VI. NH50 Area 1 16.9 9.9 4.8 7.4 0.58 0.29 0.44 142 98 51 79 0.69 0.36 0.56
Area 2 7.8 5.7 2.0 3.7 0.74 0.26 0.48 94 60 34 52 0.64 0.36 0.55
Area 3 4.9 5.0 3.2 4.7 1.01 0.66 0.96 527 208 77 116 0.39 0.15 0.22
Area 4 2.4 0.7 0.4 0.4 0.28 0.17 0.17 26 8 4 6 0.32 0.17 0.22
VII. PR15 Area 1 11.6 12.3 7.2 9.9 1.06 0.62 0.85 302 248 132 197 0.82 0.44 0.65
Area 2 6.9 6.5 4.6 5.4 0.94 0.66 0.79 45 41 26 33 0.90 0.58 0.72
Area 3 13.3 13.0 5.8 7.7 0.98 0.43 0.58 385 212 96 128 0.55 0.25 0.33
VIII. PR20 Area 1 1.9 2.0 1.4 1.6 1.05 0.74 0.85 125 110 57 87 0.88 0.46 0.69
(NH51) Area 2 8.1 6.6 4.0 5.8 0.82 0.49 0.71 146 93 51 75 0.64 0.35 0.51
Area 3 15.2 7.3 3.3 5.4 0.48 0.22 0.35 749 242 125 158 0.32 0.17 0.21
Area 4 5.9 3.3 1.7 2.3 0.56 0.28 0.39 465 71 44 44 0.15 0.09 0.09
Source: Study Team
1) Ave. occupancy: motorcycle - 1.3, car - 1.9, bus – 50
2) Toll fee on urban expressway: 15,000VND/ride; on interregional expressway: 1,000VND/km; fare on UMRT:
5,000VND/ride+500VND/km

3-28
 The Study on Urban Transport Master Plan and Feasibility Study in HCM Metropolitan Area (HOUTRANS)
Final Report
VOLUME 4: Transport Database
Figure 3.4.1 Traffic Assignment on Recommended Network: PT 30% Case
at AM Peak Hour

(
Traffic Flow
VCR<1.00
VCR<1.20
VCR<1.50
1.50<VCR

20 10 5 (000PCU)

(
Traffic Flow
VCR<1.00
VCR<1.20
VCR<1.50
1.50<VCR

20 10 5 (000PCU)

Source: Study Team

Note: Assumed modal share (%) of motorcycle, car and bus is 50, 20 and 30,
Average occupancy: motorcycle - 1.3, car - 2.0, bus – 50, respectively.

3-29
 The Study on Urban Transport Master Plan and Feasibility Study in HCM Metropolitan Area (HOUTRANS)
Final Report
VOLUME 4: Transport Database
Figure 3.4.2 Traffic Assignment on Recommended Network: PT 30% Case
at Off Peak Hour

(
Traffic Flow
VCR<1.00
VCR<1.20
VCR<1.50
1.50<VCR

20 10 5 (000PCU)

(
Traffic Flow
VCR<1.00
VCR<1.20
VCR<1.50
1.50<VCR

20 10 5 (000PCU)

Source: Study Team

Note: Assumed modal share (%) of motorcycle, car and bus is 50, 20 and 30,
Average occupancy: motorcycle - 1.3, car - 2.0, bus – 50, respectively.

3-30
 The Study on Urban Transport Master Plan and Feasibility Study in HCM Metropolitan Area (HOUTRANS)
Final Report
VOLUME 4: Transport Database
Figure 3.4.3 Traffic Assignment on Recommended Network: PT 50% Case
at AM peak hour

(
Traffic Flow
VCR<1.00
VCR<1.20
VCR<1.50
1.50<VCR

20 10 5 (000PCU)

(
Traffic Flow
VCR<1.00
VCR<1.20
VCR<1.50
1.50<VCR

20 10 5 (000PCU)

Source: Study Team

Note: Assumed modal share (%) of motorcycle, car and bus is 50, 20 and 30,
Average occupancy: motorcycle - 1.3, car - 2.0, bus – 50, respectively.

3-31
 The Study on Urban Transport Master Plan and Feasibility Study in HCM Metropolitan Area (HOUTRANS)
Final Report
VOLUME 4: Transport Database
Figure 3.4.4 Traffic Assignment on Recommended Network: PT 50% Case
at Off Peak Hour

(
Traffic Flow
VCR<1.00
VCR<1.20
VCR<1.50
1.50<VCR

20 10 5 (000PCU)

(
Traffic Flow
VCR<1.00
VCR<1.20
VCR<1.50
1.50<VCR

20 10 5 (000PCU)

Source: Study Team

Note: Assumed modal share (%) of motorcycle, car and bus is 50, 20 and 30,
Average occupancy: motorcycle - 1.3, car - 2.0, bus – 50, respectively.

3-32
 The Study on Urban Transport Master Plan and Feasibility Study in HCM Metropolitan Area (HOUTRANS)
Final Report
VOLUME 4: Transport Database
Table 3.4.3 shows the performance of networks by time period. According to this, the
performances (average travel speed and V/C ratio) vary by the time and present network
(2002) and mater plan network in 2020 shows close values on the performance.
Comparing the result of daily traffic assignment in the 2nd stage model, which was used for
the formulation and evaluation for the master plan network, the transport costs became
little bit higher because of the consideration of peak hours.

rd
Table 3.4.3 Comparison of Alternative Network Performance based on 3 Stage
Model

2020 2020/2002
2002 Do Do 1,2) Do Do
1) 1) M/P M/P
Nothing Committed Nothing Committed
Traffic Demand (mil. PCU/day) 3.1 8.2 8.2 8.2 2.6 2.6 2.6
PCU-km (mil./day) 19.0 78.1 81.3 86.1 4.1 4.3 4.5
Traffic Load
PCU-hrs (mil./day) 0.9 11.6 10.6 4.2 12.9 11.8 4.7
Ave. Travel AM-p 11.0 5.7 5.9 10.0 0.5 0.5 0.9
Travel Speed (kph) Off-p 27.5 7.0 8.2 27.7 3.0 0.3 1.0
Features Ave. V/C AM-p 0.7 2.1 1.9 1.0 0.3 2.7 1.4
Ratio Off-p 0.2 1.2 1.1 0.5 6.0 5.5 2.5
VOC 2.4 16.2 15.8 9.7 6.8 6.6 4.0
Transport
Passenger Time
Cost (mil 4.9 72.5 64.1 26.0
Cost 14.8 13.1 5.3
US$/day)
Total 7.3 88.7 79.9 35.8 12.2 4.9 10.9
Source: Study Team
1) Assumed modal share (%) of M/C, car, bus: 50/20/30, Average occupancy: M/C1.3, car2.0, bus50
2) Toll fee for Urban Expressway: 15,000VND/ride, for Interregional Expressway: 1,000VND/km and fare for
UMRT: 5,000+500/kmVND

3-33
 The Study on Urban Transport Master Plan and Feasibility Study in HCM Metropolitan Area (HOUTRANS)
Final Report
VOLUME 4: Transport Database

4 TECHNOLOGY TRANSFER

4.1 Introduction
The TDSI technical staff was involved throughout the study duration though
on-the-job-participation on the day-to-day conduct and handling of the study, thereby
learning how such studies are prepared and executed. In addition to that, there was a
specific requirement that by the end of study the local staff would be fully trained on the
use and execution of travel demand models. The model developed by the HOUTRANS
study should be converted to a suitable software familiar to the local staff and one in which
they can use with ease. This objective was based on the premise that the local staff can
use the model to test policy options, and/or develop and evaluate future transport
infrastructure projects, if so desired.
The HOUTRANS study team therefore assisted the TDSI to acquire “Trips” based
software, now called CUBE/Voyager, with which they had experience and familiarity. The
HOUTRANS study team converted the HOUTRANS STRADA-based model to the
CUBE/Voyager software. The output results of the HOUTRANS model output from the
CUBE/Voyage model were compared with the outputs of the STRADA model. This
exercise was conducted to ensure that the HOUTRANS travel demand model calibration
was valid, and the model may be run using the CUBE/Voyager software modules to
replicate the traffic forecast produced and used by the HOUTRANS study in the
development of the HCMC Master Plan.

4.2 Training Course 1 – HOUTRANS 2nd Stage Model

After the conversion of the HOUTRANS model and testing the HOUTRANS team
conducted a three day initial course to train the TDSI South staff on the use of the
HOUTRANS 2nd Stage Model. The course involved an intensive training, using the
HOUTRANS model in CUBE/Voyager software with additional examples on how to use
each sub-model. The course also trained the staff in the use of the graphics software to
develop/code new schemes for testing strategic network development options. The
course was conducted over three days. The topics covered during the each half-day long
session are summarised as follows:

Session 1: Introduction

• An Introduction to CUBE/Voyager Software,

• General discussion on its use,
• Merits of using CUBE/Voyager or any other software
• Introduction to CUBE/Voyager Graphics
• The convenience of using CUBE/Voyager File Management System

Session 2: Introduction to Demand Modelling using CUBE/Voyager programs

• Introduction to demand modeling – General points and steps/stages

• Introduction to CUBE/Voyager programs and Use for HOUTRANS 2nd Stage
Model

4-1
 The Study on Urban Transport Master Plan and Feasibility Study in HCM Metropolitan Area (HOUTRANS)
Final Report
VOLUME 4: Transport Database
• Introduction to CUBE/Voyager Setup and use of Application and Scenarios
• Transfer STRADA Networks and Trip Matrices to CUBE/Voyager formats
• Build HOUTRANS Network and Trips Matrices using CUBE/Voyager Programs

Session 3: HOUTRANS 2nd Stage Model Tripend Models in CUBE/Voyager

• Introduction to demand modeling – Trip-end Models

• 2nd Stage HOUTRANS Trip Generation Model
• 2nd Stage HOUTRANS Trip Attraction Model
• Balancing of Trip Attractions and Generations
• Use of Program GENERATION

Session 4: HOUTRANS 2nd Stage Model Trip Distribution Models in CUBE/Voyager

• Introduction to demand modeling – Trip Distribution Models

• 2nd Stage HOUTRANS Trip Distribution Models Inter and Intrazonal Models
• Implementation of HOUTRANS 2nd Stage Trip Distribution Models
• Use of Programs NETWOK, HIGHWAY to get Level of Service in terms of travel
time and Distances
• Use of Programs MATRIX for Intra and Interzonal trip distribution models.
• Analyses of results, Area to Area Movements and Trip Length Frequency
Distribution (TLD) comparison between observed and modelled trips using
program MATRIX.

Session 5: HOUTRANS 2nd Stage Model Modal Split Models in CUBE/Voyager

• Introduction to demand modeling – Modal Split Models

• 2nd Stage HOUTRANS Modal Split Models: Walk and other Mode Models
• Implementation of HOUTRANS 2nd Stage Modal Split (Walk and other Modes)
Models
• Use of Programs NETWOK, HIGHWAY to get Level of Service in terms of travel
time and Distances and costs of travel by mode of travel.
• Use of Programs MATRIX for walk mode and other mode models.
• Analyses of results, Area-to-area Movements and Trip Length Frequency
Distribution (TLD), comparison between observed and modelled trips using the
MATRIX program.

Session 6: HOUTRANS 2nd Stage Model – Traffic Assignment Model in

CUBE/Voyager

• Introduction to demand modeling – Traffic Assignment Methods

• Data Required for Traffic Assignment – Highway Capacity Calculation
• Data Required for Traffic Assignment – Highway Speed/Flow Relationships
• 2nd Stage HOUTRANS Traffic Assignment – Formation of Total Trip Matrices in
Passenger Car Units (PCUs)
• Implementation of the HOUTRANS 2nd Stage Assignment Model including the
Speed/Flow formulations to the CUBE/Voyager program HIGHWAY.

4-2
 The Study on Urban Transport Master Plan and Feasibility Study in HCM Metropolitan Area (HOUTRANS)
Final Report
VOLUME 4: Transport Database
• Use of NETWORK programs for analyses of results.
• Analyses of results compare observed and assigned traffic volume by vehicle type
and totals.
• Estimation of network-wide travel indices for evaluation.

The models used in the training are documented in the section 2 of this report. Other
subjects covered were on how to build observed trip matrices from the HOUTRANS HIS
and the Cordon data files. The names of the participants are listed in Annex 4.1.

4.3 Training Course 2 – HOUTRANS 2nd and 3rd Stage Models

After development of the 2nd Stage HOUTRANS 2020 model on the CUBE/Voyager
software an additional training course which lasted six (6) days were held in HCMC. The
course was attended by both the TDSI South and North team members. The abstract of
the contents of the course are shown in Annex 4.2 along with the list of participants. The
course was conducted with the HOUTRANS 2nd Stage model used in the TDSI staff
computers, along with hands-on sessions.

By the end of the 2nd course the TDSI staff showed confidence that they would be able to
run the model as taught to them. The duration of the course training stressed the how-tos
of interpreting model outputs, carrying out simple checks to ensure that the modelled
forecast are meaningful, sensible, and could be used to carry out economic, financial and
environmental evaluation of the options being tested. The tables at the end of this chapter
summarize the subjects/topics covered by the six-day training course.

4.4 Evaluation and Comments on Training Courses

Table 4.4.1 shows the evaluation on the trainees. Evaluations on the lectures and the
CUBE/Voyager were generally high. The trainees also had a high evaluation of the
CUBE/Voyager itself. However, understanding of the CUBE/Voyager comparatively had
lower assessment, thereby necessitating a longer and slower lecture, which they
requested.

4-3
 The Study on Urban Transport Master Plan and Feasibility Study in HCM Metropolitan Area (HOUTRANS)
Final Report
VOLUME 4: Transport Database
Table 4.4.1 Evaluation of the Training

Total TDSI - South Others

Item Bad & Very Bad & Very Bad & Very
So- So- So-
Very Good & Very Good & Very Good &
so so so
Bad Good Bad Good Bad Good
Assessment of Lecture

1. Level of Lecture - 4 3 - 2 1 - 2 2

2. Explanation 2 1 4 1 1 1 1 - 3

3. Handout 2 4 1 2 1 - - 3 1
Level of Understanding

1. Demand Forecast Theory - 4 3 - 2 1 - 2 2

2. Model Equations - 5 2 - 2 1 - 3 1

3. Model Building - 6 1 - 2 1 - 4 -

4. Model Parameters 1 4 2 - 2 1 1 2 1

5. HOUTRANS Model - 6 1 - 2 1 - 4 -

6. How to use CUBE/Voyager - 7 - - 3 - - 4 -

Assessment of CUBE/Voyager

1. Integrity - 3 4 - 2 1 - 1 3

2. User Interface - 2 5 - - 3 - 2 2

3. Graphics - 3 4 - - 3 - 3 1

4. Model Options - 1 6 - - 3 - 1 3

5. Applicability to Use in TDSI - 1 6 - 1 2 - - 4

Source: Study Team

4-4
 The Study on Urban Transport Master Plan and Feasibility Study in HCM Metropolitan Area (HOUTRANS)
Final Report
VOLUME 4: Transport Database
The TDSI staff who attended the courses were enthusiastic about learning new skills and
in being able to improve their current knowledge on demand modeling and the use of
Trips’ software. It was stressed throughout the course that the modeling process is only an
estimation process and a great deal of effort, skill and experience would be required to
carry out the modeling work. They staff were taught numerous examples on how to
analyze results and how to conduct simple “what-if checks” on the model outputs to
ensure ‘sensible’ results that may be used for further analyses.
During the course some trainees showed difficulty in understanding hard-to-grasp
terminologies in transport modeling, especially the language used in the CUBE/Voyager
Software. In order to overcome the language difficulties the staff were repeatedly
requested to interrupt at any time whenever they encountered difficult terminologies or
words. On many occasions the senior staff, especially those whose English were better
were asked to further explain terminologies and procedures in Vietnamese to the younger
or less experienced staffers.

They were also taught repeatedly in the use of the CUBE/Voyager “Help” so as to acquaint
them on the available options and choices in each program. It was stressed to them that
they could only improve their skills by constant “practice” and continuous use of the
software.

A few weeks after the training they were given questionnaire forms for feed backing so as
to ascertain how the TDSI staff were getting on in the use of the HOUTRANS model, the
problems and difficulties they were facing, and as to what other information, lessons or
training would be more useful to them? The results of the questionnaire are summarized
in the next section.

The opinion of the HOUTRANS Team is that what is lacking is a “continuous” hands-on
use of the models and the use of various model runs to gain the necessary experience.
However, it was noted that the staffs were capable of understanding the problems, albeit
they still lacked some confidence in solving them. It is stressed that “confidence” comes
with experience.

4-5
 The Study on Urban Transport Master Plan and Feasibility Study in HCM Metropolitan Area (HOUTRANS)
Final Report
VOLUME 4: Transport Database
Table 4.4.2 Training Schedule

Friday 13th February 2004 AM – Session-1

Introduction and Overview of Transport Planning Process

Topics Covered:
• Transport Planning process in the context of Planning
• Principles of Transport Planning
• Travel Demand forecasts – Why forecast?
• Travel Demand forecasting – Type of models
• Travel Demand surveys
• Development, calibration and validation of Travel Demand models
• Forecasting Travel Demands for alternatives and policy tests
• Evaluation of Alternatives

AM – Session-2

Travel Demand Surveys – HOUTRANS Surveys and Experience

Topics Covered:
• Types of HOUTRANS travel surveys conducted
• HOUTRANS HIS:
– Purpose, design and conduct of surveys,
– Problems and issues during the HIS,
– Analyses of the HIS data and outputs.
– Development of base year trip matrices,
• Roadside Interview Surveys (RIS):
– Purpose, design and conduct of surveys,
– Problems and issues during RIS,
– Analysis of the RIS data and outputs.
– Development of trip matrices.
• Traffic Count Surveys at Study Area Cordon and Screen-lines;
• Other Travel Demand Surveys:
– Journey Time Surveys
– Bus Passenger Surveys
– Special Terminal Surveys at Ports & Bus Termini,
– Stated Preference Surveys
– Highway network inventory
– Other associated surveys…

Lunch Break

4-6
 The Study on Urban Transport Master Plan and Feasibility Study in HCM Metropolitan Area (HOUTRANS)
Final Report
VOLUME 4: Transport Database
Table 4.4.2 Training Schedule (continued)

Friday 13th February 2004 PM – Session-1

Travel Demand Modeling – Types of Models

Topics Covered:
• Development of Multi-Stage HOUTRANS Travel Demand Models:
Description, pros and cons and the use of each type model.
• Stage-1 HOUTRANS Land use Forecast Model:
– Model development and definition
– Model inputs and level of details (district level data and corridors for traffic)
– Model used to test alternative land use development scenarios
• Stage-2 HOUTRANS Strategic (Aggregate) Demand Forecast Model:
– Model development and definition,
– Model inputs and level of details (traffic zone level data and main roads of
the study area,
– Model used to test alternative Transport Master Plan development
scenarios
• Stage-3 HOUTRANS Dis-aggregate Demand Forecast Model
– Model development and definition by household categories and by trip
purpose,
– Model inputs and level of details (traffic zone level data and main roads of
the study area, forecasts of HH types by vehicle ownership and
availability)
– Model used to test alternative transport policies and scenarios.
– Later adopted to test alternatives at feasibility level.

PM – Session-2

Travel Demand Modeling – HOUTRANS Stage-2 Demand Forecast Model

Topics Covered:
• Application of 4-Step demand modeling process for HOUTRANS Stage-2
Model: Brief definition of each step and its application:
• Step-1: Trip Generation & Attraction Models
• Step-2: Trip Distribution Model
• Step-3: Modal Split Model
• Step-4: Traffic Assignment Model
• Evaluation of Results

End of Day Summary and Discussion, Further Q&A.

4-7
 The Study on Urban Transport Master Plan and Feasibility Study in HCM Metropolitan Area (HOUTRANS)
Final Report
VOLUME 4: Transport Database
Table 4.4.2 Training Schedule (continued)

Monday 16th February 2004 AM – Session-1

Introduction and Overview of CUBE/Voyager Software for Travel Demand

Modeling

Topics Covered:
• Types of software available for travel demand forecasting
• Pros and cons of different types of software packages
• Initial use of STRADA for the development and application of HOUTRANS
multi-stage models
• Transfer of HOUTRANS Stage-2 Model to the CUBE/Voyager software
• Introduction to the CUBE/Voyager software
• Hands on experience in use and setup of modeling environment in the CUBE
software

AM – Session-2

HOUTRANS Stage-2 Model Highway Network

Topics Covered:
• Development of HOUTRANS study area highway network;
– Purpose and usage of highway network in travel demand modeling,
– Data collection for highway network; i.e. Highway network inventory,
– Attributes collected for highway network model,
– Calculation of highway capacity,
– Development of speed/flow (QV) relationships
• Representation of highway network using CUBE/Voyager
– Building of Highway network using Voyager Program, NETWORK:
– Data requirements of the program
– Program outputs and usage of output file(s)
– Viewing and Manipulation of network file using VIPER Program.
– Development of trip matrices.
• Hands on experience in the use of Programs NETWORK & VIPER

Lunch Break

4-8
 The Study on Urban Transport Master Plan and Feasibility Study in HCM Metropolitan Area (HOUTRANS)
Final Report
VOLUME 4: Transport Database
Table 4.4.2 Training Schedule (continued)

Monday 16th February 2004 PM – Session-1

Development of Trip Matrices and Matrix Manipulations

Topics Covered:
• How “travel-trips” are represented using a “Matrix”
• Sources of data for building of trip matrices, i.e. HIS, RIS or direct interview
surveys on buses or at ports and other transport termini.
• Types of software used for manipulation of survey data: database programs
or CUBE/Voyager Program MATRIX.

PM – Session-2

Building HOUTRANS Observed Trip Matrices

Topics Covered:
• Defining the data input for building trip matrices,
• Introduction to CUBE/Voyager program for building trip matrices, MATRIX.
• Building “observed” trip matrices using Program MATRIX from HOUTRANS
HIS and RIS data files.
• Matrix calculations and obtaining new matrices.
• Compression of zonal trip matrices to district or sector level for easy
assessment.
• Viewing and obtaining matrix “statistics” using Program VIPER.
• Addition, subtraction and matrix manipulations using MATRIX Programs
• Hands on experience in the use and running of Program MATRIX by staff to
compute and manipulate trip matrices.
• Hands on experience in the use of Programs MATRIX & VIPER, for matrix
manipulations, viewing and obtaining statistics.

End of Day Summary and Discussion, Further Q&A.

4-9
 The Study on Urban Transport Master Plan and Feasibility Study in HCM Metropolitan Area (HOUTRANS)
Final Report
VOLUME 4: Transport Database
Table 4.4.2 Training Schedule (continued)

Tuesday 17th February 2004 AM – Session-1

Introduction and Overview of Trip-end (Trip Generation and Attraction) Models

for HOUTRANS Stage-2 Model

Topics Covered:
• Trip Generation models types and description.
• Development of Stage-2 trip generation models and “statistically significant”
parameters.
• Development of Stage-2 trip attraction models and “statistically significant”
parameters.
• Data availability for base year trip generation and attraction model variable
data (e.g. Planning Data on Population, Households, Jobs, School places etc.
• Forecast of Trip-end model planning data for forecast years
• Methods of forecasting future year planning data variables used by the
trip-end models.

AM – Session-2

HOUTRANS Stage-2 Trip-end Models in CUBE/Voyager Software

Topics Covered:
• Development of HOUTRANS Stage-2 trip generation and attraction models
using Program GENERATION.
• Preparation of planning data for study area for trip-end models.
• Balancing of generation and attractions to each other and to exogenously
forecast control totals.
• Running Program GENERATION and interpretation of results;
• Hands on experience in the use of Program GENERATION and analysis of
output.

Lunch Break

4-10
 The Study on Urban Transport Master Plan and Feasibility Study in HCM Metropolitan Area (HOUTRANS)
Final Report
VOLUME 4: Transport Database
Table 4.4.2 Training Schedule (continued)

Tuesday 17th February 2004 PM – Session-1

Introduction to Trip Distribution Models, and HOUTRANS Stage-2 Trip

Distribution Models

Topics Covered:
• Types of trip distribution models in transport demand forecasting – an
overview
• Trip distribution models used in HOUTRANS stage-2;
• Calibration and validation of trip distribution models.
• Sources of data for trip distribution models: e.g. trip-ends, zone-to-zone
impedance values.
• Building of impedance matrices from highway network in CUBE/Voyager.
• Conversion of HOUTRANS Stage-2 trip distribution models based on
STRADA software to CUBE/Voyager.

PM – Session-2

Building HOUTRANS Stage-2 Trip Distribution models

Topics Covered:
• Defining the data input for Intrazonal trip distribution model.
• Implementation of intrazonal trip distribution model using Program MATRIX.
• Building of impedance matrix for Interzonal trip distribution model,
comparable to those used by STRADA.
• Implementation of interzonal trip distribution model using Program MATRIX.
• Compression of interzonal trip matrix to district or sector level for comparison
with observed trip matrix.
• Computing Trip Length frequency Distribution (TLD) using Program MATRIX.
• Comparing the “observed” and “modeled” TLD for observed and modeled
distribution patterns.
• Hands on experience in the use and running of Program MATRIX by staff for
trip distribution models.

End of Day Summary and Discussion, Further Q&A.

4-11
 The Study on Urban Transport Master Plan and Feasibility Study in HCM Metropolitan Area (HOUTRANS)
Final Report
VOLUME 4: Transport Database
Table 4.4.2 Training Schedule (continued)

Wednesday 18th February 2004 AM – Session-1

Review of Trip Distribution Models – from Yesterday &

Introduction to Modal Split Models

Topics Covered:
• Trip distribution model calibration and validation methods,
• Review of trip distribution model results.
• Assessment of calibration results – goodness of fit of the trip distribution
models.
• Introduction to Modal Split Models.
• Description of HOUTRANS Stage-2 Modal Split model.
• Definition of Stage-2 2-Step modal split model
• 1) Description of Walk Trip modal split.
• 2) Description of Non-Walk (mechanized) mode modal split model, using
Logit model.
• Attributes of data required on the cost of travel for by each mode of travel for
the modal split model.

AM – Session-2

HOUTRANS Stage-2 Walk Modal Split Model Under CUBE/Voyager

Topics Covered:
• Conversion of HOUTRANS Stage-2 STRADA based Modal Split Model to
CUBE/Voyager;
• Calculation of travel distance using Voyager Programs HIGHWAY and
MATRIX.
• Calculation of Walk modal split matrix from total matrix in Voyager
• Comparison of results between observed and modeled trips by comparing:
– Trips at district and sector levels;
– Comparing observed and modeled trip lengths.
– General discussion on the results and merits of HOUTRANS Stage-2 modal
split model.
• Hands on experience in the use of Programs HIGHWAY & MATRIX for use in
modal split models.

Lunch Break

4-12
 The Study on Urban Transport Master Plan and Feasibility Study in HCM Metropolitan Area (HOUTRANS)
Final Report
VOLUME 4: Transport Database
Table 4.4.2 Training Schedule (continued)

Wednesday 18th February 2004 PM – Session-1

Review of Trip Modal Split Models – from AM Sessions &

Introduction to Multi-mode Modal Split (M/S) Models

Topics Covered:
• Skimming of highway travel distances and times for M/S model using
program HIGHWAY.
• Calculation for cost by each mode for use in modal split models using
Program MATRIX.
• Calculation of distance based fares for input to modal split models.
• Representation of Stage-2 multi-mode STRADA based M/S model using
Voyage Program MATRIX.

PM – Session-2

HOUTRANS Stage-2 ALL Modal Split Models Under CUBE/Voyager

Topics Covered:
• Comparison of results between observed and modeled model split models by
comparing:
−Trips at district and sector levels;
−Comparing observed and modeled trip lengths.

• General discussion on the results and merits of HOUTRANS Stage-2 modal

split model
• Hands on experience in the use of Programs HIGHWAY & MATRIX for use in
modal split models.

End of Day Summary and Discussion, Further Q&A.

4-13
 The Study on Urban Transport Master Plan and Feasibility Study in HCM Metropolitan Area (HOUTRANS)
Final Report
VOLUME 4: Transport Database
Table 4.4.2 Training Schedule (continued)

Thursday 19th February 2004 AM – Session-1

Introduction to Traffic Assignment Model

Topics Covered:
• Introduction to traffic assignment, its purpose and need.
• Review of traffic assignment techniques and theory behind each technique.
• Merits of different traffic assignment techniques (All-or-Nothing, Capacity
restraint, Volume Averaging, Incremental Assignment & Equilibrium
Assignment) and methodologies and their availability in different software
packages (e.g. STRADA, Voyager, Emme/2, Saturn & Paramics, etc)
• Review of types of software and models available for traffic assignment
process.

AM – Session-2

HOUTRANS Stage-2 Traffic Assignment Model

Topics Covered:
• Conversion of modal split model output person trip matrices to PCU trip
matrices for traffic assignment model.
• Assessment and review of highway link capacities used in the traffic
assignment process.
• Development of Speed/Flow (S/F) relationships for capacity restraint traffic
assignment for Stage-2 models.
• Discussion on the STRADA model capacity restraint assignment results using
different S/F relationships.
• Assessment of “daily” traffic assignment method (incremental assignment)
used for Stage-2 model using STRADA.
• Assessment of STRADA traffic assignment results.

Lunch Break

4-14
 The Study on Urban Transport Master Plan and Feasibility Study in HCM Metropolitan Area (HOUTRANS)
Final Report
VOLUME 4: Transport Database
Table 4.4.2 Training Schedule (continued)

Thursday 19th February 2004 PM – Session-1

Introduction to Traffic Assignment Model – Under Voyager

Program HIGHWAY.

Topics Covered:
• Coding the choice of method to use for traffic assignment
(incremental) in Program HIGHWAY.
• Techniques used to represent STRADA S/F relationship in
Program HIGHWAY.
• Running of Program HIGHWAY and saving of assignment
results for analysis later.

PM – Session-2

Traffic Assignment Model Under Voyager & Analysis of

Results

Topics Covered:
• Review of Assignment results and interpretation
• Analysis of results from each iteration of the assignment
process using Program NETWORK.
• Analysis of traffic assignment – traffic volumes by screen
lines, and comparison with observed traffic volumes.
• Calculation of network-wide journey times and comparison
of results with observed journey times.
• Analysis of assignment results from output database file
from Program HIGHWAY.

End of Day Summary and Discussion, Further Q&A.

4-15
 The Study on Urban Transport Master Plan and Feasibility Study in HCM Metropolitan Area (HOUTRANS)
Final Report
VOLUME 4: Transport Database
Table 4.4.2 Training Schedule (continued)

Friday 20th February 2004 AM – Session-1

Review of the Highway Modeling Process and General Q&A.

Topics Covered:
• Prepare 2020 model Run inputs.
• Hands on 2020 Model run by TDSI staff and Q&A.
• Interpretation of results and assessment of results.
• Discussion on evaluation of model outputs.

AM – Session-2

General Discussion on HOUTRANS Stage-3 Model and its Merits.

Topics Covered:
• Advantages and disadvantages of disaggregate models.
• Application of HOUTRANS Stage-3 model to test policy options –an overview

Friday 20th February 2004 PM – Session-1

General Discussion on Public Transport (PT) Assignment Using Voyager

Topics Covered:
• Summary of data required by public transport assignment program of
Voyager PUBLIC TRANSPOT e.g. Highway network, public transport route
data input as line data, frequency of operation, other attributes of PT system
fares, etc.
• Coding of line data for Program PUBLIC TRANSPOT.
• Building of PT network.
• Viewing of line data.
• PT Assignment program not yet fully implemented under voyager, but
covered its principles and methodology.

End of Day Summary and Discussion, Further Q&A and Discussion

4-16
 The Study on Urban Transport Master Plan and Feasibility Study in HCM Metropolitan Area (HOUTRANS)
Final Report
VOLUME 4: Transport Database

ANNEX 1 HOUTRANS Study Area Zone System Definition

The following Table defines the Houyrans Study Area Zone System and its relation to local
political/administrative districts.
Table 1.1 Study Area Zone System Definition by District

District District Start End

Zones Area
Number Name Zone Zone
1 1 1 12 12 1
2 2 137 141 5 3
3 3 13 24 12 1
4 4 25 32 8 1
5 5 33 47 15 1
6 6 48 59 12 1
7 7 142 147 6 3
8 8 60 72 13 2
9 9 148 156 9 3
10 10 73 84 12 1
11 11 85 97 13 1
12 12 157 163 7 3
13 Go Vap 98 105 8 2
14 Tan Binh 114 129 16 2
15 Binh Thanh 106 113 8 2
16 Phu Nhuan 130 136 7 1
17 Thu Duc 201 209 9 3
18 Hoc Mon 191 200 10 4
19 Binh Chanh 171 190 20 3
20 Nha Be 164 170 7 4
21 Cu Chi 210 215 6 5
22 Can Gio 216 216 1 5
23 Thu Dau Mot 240 243 4 6
24 Thuan An 244 245 2 7
25 Di An 246 247 2 7
26 Bien Hoa 248 258 11 6
27 Long Thanh 259 262 4 8
28 Nhon Trach 263 265 3 8
29 Tan An 217 218 2 6
30 Can Giuoc 219 222 4 8
31 Can Duoc 223 226 4 8
32 Chau Thanh 227 228 2 8
33 Thu Thua 229 230 2 8
34 Ben Luc 231 233 3 8
35 Duc Hoa 234 238 5 8
36 Tan Tru 239 239 1 8

A1-1
 The Study on Urban Transport Master Plan and Feasibility Study in HCM Metropolitan Area (HOUTRANS)
Final Report
VOLUME 4: Transport Database

Table 1.2 Houtrans Study Area Definition of Areas for Analyses

Area Area Name Districts Zones Zones Zones

HCMC CBD 1, 3, 5, 10 1-24 33-47 73-84
1
HCMC Inner Area 4, 6, 11, 16 25-32 48-59 85-97 &130-136
2 HCMC Inner Fringe 8, 13-15 60-72 98-129
3 HCMC Peripheral 2, 7, 9, 12, 17, 19 137-163 171-190 201-209
4 HCMC Suburban 18, 20 164-170 191-200
5 HCMC Rural 21-22 210-216
6 Satellite Urban 23, 26, 29 217-218 240-243 248-258
7 Suburban 24-25 244-247
8 Rural 27-28, 30-36 219-239 259-265
Externals - 266-270

A1-2
 The Study on Urban Transport Master Plan and Feasibility Study in HCM Metropolitan Area (HOUTRANS)
Final Report
VOLUME 4: Transport Database

ANNEX 2 CUBE/VOYAGER MODEL DESCRIPTION (1)

2.1 BUILD AND SKIM NETWORKS

The CUBE/Voyager Model is shown below. The base case network is built by converting the
Strada model network to two DBF files: 1. Link data; 2. Node Data. Then the binary network file is
used to build Time & Distance Skim matrices.

Similarly build the highway network with peak capacities.

1. Program: NETWORK – Builds the Binary Network file – Using 24-Hour Capacities
Input Files:
1. Link Data File - Converted from Strada Model Using XL Spreadsheet
2. Node Data File - Converted from Strada Model Using XL Spreadsheet
Output File:
1. Binary Network File.

2. Program: HIGHWAY – Builds the TIME and DISTANCE Skim Matrices

Input Files:
1. Binary Network file from previous program.
Output File:
2. MATRIX File containing Time & Distance Skims

3. Program: NETWORK – Builds Binary Network – Using Hourly Capacities

Input Files:
1. Binary Network file from previous program.
Output File:
2. MATRIX File containing Time & Distance Skims

Script Files are Given on the following Pages:

A2-1
 The Study on Urban Transport Master Plan and Feasibility Study in HCM Metropolitan Area (HOUTRANS)
Final Report
VOLUME 4: Transport Database

Script File 1.
Do not change filenames or add or remove FILEI/FILEO statements using an editor.
Use Cube/Application Manager.
RUN PGM=NETWORK PRNFILE="C:\HOUTRANS\BASE02\BNNET00A.PRN" MSG='24_HR_Build
Network'
FILEO NETO = "C:\HOUTRANS\BASE02\BNNET00A.NET"
FILEI NODEI[1] = "C:\HOUTRANS\BASE02\INPUT_DATA\NODE_2002-V2.DBF"
FILEI LINKI[1] = "C:\HOUTRANS\BASE02\INPUT_DATA\LINK_2002-V2.DBF"
;
; Build Network from Link & Node data File
;
Zones=270
;
;
;
FF_TIME=INT((LI.1.DISTANCE/LI.1.SPEED)*6000)*.01
IF(FF_TIME<0.01) FF_TIME=0.01
;
;
ENDRUN

Script File 2.
; Do not change filenames or add or remove FILEI/FILEO statements using an editor.
Use Cube/Application Manager.
RUN PGM=HIGHWAY PRNFILE="C:\HOUTRANS\BASE02\BNHWY00A.PRN" MSG='Build Time &
Dist Skims'
FILEI NETI = "C:\HOUTRANS\BASE02\BNNET00A.NET"
;
FILEO MATO[1] = "C:\HOUTRANS\BASE02\TIME&DIST.MAT", Dec=2,
MO=1-2 NAME=TIME, DIST
; Read in Network
PROCESS PHASE=LINKREAD
LINKCLASS=LI.CLASS
ENDPHASE
; Build and skim Path Loop for all zones
;
PROCESS PHASE=ILOOP
PATHLOAD PATH=DIST,
MW[1]=PATHTRACE(TIME), MW[2]=PATHTRACE(LI.DISTANCE)
MW[1][I]=0 ; Set Intrazonal Values to Zero
MW[2][I]=0
ENDPHASE
;
ENDRUN

Script File 3.
; Do not change filenames or add or remove FILEI/FILEO statements using an editor.
Use Cube/Application Manager.
RUN PGM=NETWORK PRNFILE="C:\HOUTRANS\BASE02\BNNET00B.PRN" MSG='Build Peak
Network = Peak Capacities'
FILEI NODEI[1] = "C:\HOUTRANS\BASE02\INPUT_DATA\NODE02_PK.DBF"
FILEI LINKI[1] = "C:\HOUTRANS\BASE02\INPUT_DATA\LINK02_PK.DBF"
FILEO NETO = "C:\HOUTRANS\BASE02\INPUT_DATA\BY02_PEAK.NET"
;
Zones=270
;
ENDRUN

A2-2
 The Study on Urban Transport Master Plan and Feasibility Study in HCM Metropolitan Area (HOUTRANS)
Final Report
VOLUME 4: Transport Database

2.2 BUILD OBSERVED TRIP MATRICES

The CUBE/Voyager Model is shown below.

Program Used: MATRIX

Input Files:
1. HIS Survey Data File

2. Cordon Survey Data File

Output Files:
1. Observed Total (Internal + Cordon Crossing)Person Trip Matrices File.

2. Observed (Internal + Cordon Crossing)PCU Trip Matrices File.

Script File is Given on the Next Page:

A2-3
 The Study on Urban Transport Master Plan and Feasibility Study in HCM Metropolitan Area (HOUTRANS)
Final Report
VOLUME 4: Transport Database

; Do not change filenames or add or remove FILEI/FILEO statements using an editor.

Use Cube/Application Manager.
RUN PGM=MATRIX PRNFILE="C:\HOUTRANS\BASE02\BMMAT00C.PRN" MSG='Build Obs Trip
Mxs'
; Observed Survey Data Files in IJM:V Pattern
FILEI MATI[1] = "C:\HOUTRANS\BASE02\INPUT_DATA\HIS_6_MODES.DAT",
PATTERN=IJM:V,FIELDS=1-3,4-6,7-8,9-14
;
FILEI MATI[2] = "C:\HOUTRANS\BASE02\INPUT_DATA\CRD_6_MODES.DAT",
PATTERN=IJM:V,FIELDS=1-3,4-6,7-8,9-14;
; Output Matrix
FILEO MATO[1] = "C:\HOUTRANS\BASE02\INPUT_DATA\OBS_PRSN_TRIPS_02.MAT",
MO=1-8, NAME=WALK,BCY,MCY,CAR,TRUCK,PT,Tot_Prsn, Indx
;
FILEO MATO[2] = "C:\HOUTRANS\BASE02\INPUT_DATA\OBS_PCU_TRIPS_02.MAT",
MO=11-17 Name=BCY_PCU, MCY_PCU, Car_PCU, TRUCK_PCU, PT_PCU, Tot_PCU, Indx
;
ZONES=270
DEC=2
;
; Build Person Trips Mxs
;
MW[1]=MI.1.1+MI.2.1 ; WALK (HIS+CORDON)
MW[2]=MI.1.2+MI.2.2 ; BCY (HIS+CORDON)
MW[3]=MI.1.3+MI.2.3 ; MCY (HIS+CORDON)
MW[4]=MI.1.4+MI.2.4 ; CAR (HIS+CORDON)
MW[5]=MI.1.5+MI.2.5 ; TRUCK (HIS+CORDON)
MW[6]=MI.1.6+MI.2.6 ; PT (HIS+CORDON)
;
MW[7]=MW[1]+MW[2]+MW[3]+MW[4]+MW[5]+MW[6]
MW[8]=0
;
JLOOP
IF (MW[7][J]>0) MW[8][J]=1 ; Create Matrix=1 if Cell Value>0 for Person Trips.
ENDJLOOP
;
; Build PCU Mxs
;
MW[11] = MW[2] * 0.2/ 1.14 ; BCY PCU= OCC=1.14, PCU_Fact=0.2
MW[12] = MW[3] * 0.2/ 1.51 ; MCY PCU= OCC=1.51, PCU_Fact=0.2
;
MW[13] = MW[4] * 1.0/ 1.96 ; CAR PCU= OCC=1.96, PCU_Fact=1.0
MW[14] = MW[5] * 2.5/ 1.98 ; TRUCK PCU= OCC=1.98, PCU_Fact=2.5
MW[15] = MW[6] * 2.0/ 8.18 ; PT PCU= OCC=8.18, PCU_Fact=2.0
;
MW[16]=MW[11]+MW[12]+MW[13]+MW[14]+MW[15]
MW[17]=0
;
JLOOP
IF (MW[16][J]>0) MW[17][J]=1 ; Create Matrix=1 if Cell Value>0 for PCU Matrix
ENDJLOOP
;
;
ENDRUN

A2-4
 The Study on Urban Transport Master Plan and Feasibility Study in HCM Metropolitan Area (HOUTRANS)
Final Report
VOLUME 4: Transport Database

2.3 TRIP GENERATION AND ATTRACTION MODEL

The CUBE/Voyager Model is shown below. The base year person tripends are built in a single
program run of “GENERATIONS”.

Program Used: GENERATIONS

Input Files:

1. Base Year Planning Data File

2. Forecast Year Planning Data File (In case of Base Year Run use Base Year file gaain)

Output Files:

1. Tripends and Growth Factors File

Notes:
In the base case model runs the growth factors are set to 1.0. However, in a forecast year run the
growth factors are estimated according to the growth in Population, School places and Jobs.

Script File is Given Below:

; Do not change filenames or add or remove FILEI/FILEO statements using an editor.
Use Cube/Application Manager.
RUN PGM=GENERATION PRNFILE="C:\HOUTRANS\BASE02\BTGEN00A.PRN" MSG='Estimate
Int Trip_Ends & Growth Factors for Ext'
FILEI ZDATI[2] = "C:\HOUTRANS\BASE02\INPUT_DATA\PD_2002.DBF"; Base Year Data
File
;
FILEI ZDATI[1] = "C:\HOUTRANS\BASE02\INPUT_DATA\PD_2002.DBF"; Forecast Year
Planning Data
;
;
FILEO PAO[1] = "C:\HOUTRANS\BASE02\BTGEN00A.DBF",
DBF=T,
LIST=Z(4.), P[1](8.0), A[1](8.0), P[2](6.4), A[2](6.4)
;

A2-5
 The Study on Urban Transport Master Plan and Feasibility Study in HCM Metropolitan Area (HOUTRANS)
Final Report
VOLUME 4: Transport Database

;
PARAMETERS ZONES=270
;
PROCESS PHASE=ILOOP ; Calculate Trip_Ends
& Factors
;
Pop_G = 1.404
WRK_G = 2.627
SCL_G = 2.284
G_Con = -1054.0
;
Pop_A = 1.429
WRK_A = 2.524
SCL_A = 2.333
A_Con = -869.5
;
POP_FY = ZI.1.POP
WRK_FY = ZI.1.WRK
SCL_FY = ZI.1.SCL
;
POP_BY = ZI.2.POP
WRK_BY = ZI.2.WRK
SCL_BY = ZI.2.SCL
;
; Calc. Generations
;
P[1] = Pop_G * POP_FY + WRK_G * WRK_FY + SCL_G * SCL_FY + G_Con
IF( P[1]<=0 ) P[1]=0.0 ; In case Silly Plan Data as G_Con<0.0, TripEnds could
be Negative
;
; Calc. Attractions
;
A[1] = Pop_A * POP_FY + WRK_A * WRK_FY + SCL_A * SCL_FY + A_Con
IF( A[1]<=0 ) A[1]=0.0; In case Silly Plan Data as A_Con<0.0, TripEnds could
be Negative
;
; End of Calculation of External Trip_Growth Factors
;
TPOP_BY = TPOP_BY + POP_BY
TWRK_BY = TWRK_BY + WRK_BY
TSCL_BY = TSCL_BY + SCL_BY
;
TPOP_FY = TPOP_FY + POP_FY
TWRK_FY = TWRK_FY + WRK_FY
TSCL_FY = TSCL_FY + SCL_FY
;
P[2]=1.0
A[2]=1.0
IF ( POP_BY>0.0) P[2]=POP_FY/POP_BY
IF ( (WRK_BY+SCL_by) > 0.0) A[2]=(WRK_FY+SCL_FY)/(WRK_BY+SCL_BY)
;
; Adjust Atts=Gens
;
ENDPHASE
;
PROCESS PHASE=ADJUST
;
;
P[1][266]= 0; Gen[266]
P[1][267]= 0; Gen[267]
P[1][268]= 0; Gen[268]

A2-6
 The Study on Urban Transport Master Plan and Feasibility Study in HCM Metropolitan Area (HOUTRANS)
Final Report
VOLUME 4: Transport Database

P[1][269]= 0; Gen[269]
P[1][270]= 0; Gen[270]

;
A[1][266]= 0; ATT[266]
A[1][267]= 0; ATT[267]
A[1][268]= 0; ATT[268]
A[1][269]= 0; ATT[269]
A[1][270]= 0; ATT[270]
;
P[2][266]=TPOP_FY / TPOP_BY
P[2][267]=TPOP_FY / TPOP_BY
P[2][268]=TPOP_FY / TPOP_BY
P[2][269]=TPOP_FY / TPOP_BY
P[2][270]=TPOP_FY / TPOP_BY
;
A[2][266]=(TWRK_FY + TSCL_FY) / (TWRK_BY + TSCL_BY)
A[2][267]=(TWRK_FY + TSCL_FY) / (TWRK_BY + TSCL_BY)
A[2][268]=(TWRK_FY + TSCL_FY) / (TWRK_BY + TSCL_BY)
A[2][269]=(TWRK_FY + TSCL_FY) / (TWRK_BY + TSCL_BY)
A[2][270]=(TWRK_FY + TSCL_FY) / (TWRK_BY + TSCL_BY)
;
;
BALANCE A2P=1 ; Set Att_Total to Pro_Totals 1-Set
;
ENDPHASE
;
;
ENDRUN

A2-7
 The Study on Urban Transport Master Plan and Feasibility Study in HCM Metropolitan Area (HOUTRANS)
Final Report
VOLUME 4: Transport Database

2.4 TRIP DISTRIBUTION MODEL

The Trip Distribution in CUBE/Voyager Software is carried out in three steps as shown below.
These are:

1. Distribute Intrazonal and Interzonal (Internal (zones 265) x Internal (zones 265);

2. Estimate (Internal <> External) & Exyernal <> External Trips and add to Matrix from
Stage-1 Above.
3. Control Total Number of Distributed Trips to Overall Expected Growth.

The Other two programs with execution order set to Zero(0) are for analysis only.

1. Distribute Internal (zones 1-265) Internal Trips

Program Used: MATRIX

Input Files:

1. Base Year Observed Trips Matrix and Observed Cells Index File;
2. Time & Distance Skim File;
3. Tripends as P1 & A1 and Growth Factors as P2 & A2 File, Output by the Tripend Model;

4. File containing Proportion of Intrazonals as DBF File; and

5. File containing the Centroid Connector Lengths.

Output File:

A2-8
 The Study on Urban Transport Master Plan and Feasibility Study in HCM Metropolitan Area (HOUTRANS)
Final Report
VOLUME 4: Transport Database

1. Out Internal <> Internal Trip Matrix Including Intrazonal Trips.

2. Distribute Internal <> External and External<>External Trips

Program Used: MATRIX

Input Files:
1. Base Year Observed Trips Matrix and Observed Cells Index File;

2. Internal<>Internal Distributed Trip Matrix Output by Step-1; and

3. Tripends as P1 & A1 and Growth Factors as P2 & A2 File, Output by the Tripend Model;

Output File:

1. Full Distributed Trip Matrix.

3. Control Number of Total Distributed Trips

Program Used: MATRIX

Input Files:

1. Base Year Observed Trips Matrix and Observed Cells Index File;
2. Distributed Trip Matrix Output by Step-2; and

3. Tripends as P1 & A1 and Growth Factors as P2 & A2 File, Output by the Tripend Model;

Output File:
1. Full Distributed Trip Matrix

Script Files for 3-Steps are Given on the Following Pages:

A2-9
 The Study on Urban Transport Master Plan and Feasibility Study in HCM Metropolitan Area (HOUTRANS)
Final Report
VOLUME 4: Transport Database

1. Distribute Internal (zones 1-265) Internal Trips Script File

; Do not change filenames or add or remove FILEI/FILEO statements using an editor.
Use Cube/Application Manager.
RUN PGM=MATRIX PRNFILE="C:\HOUTRANS\BASE02\BDMAT00A.PRN" MSG='Base Year Dist
(IntxInt) Trips'
FILEI ZDATI[3] = "C:\HOUTRANS\BASE02\INPUT_DATA\INTRA_DIST.DBF"
;
FILEI MATI[1] = "C:\HOUTRANS\BASE02\INPUT_DATA\OBS_PRSN_TRIPS_02.MAT"
FILEI MATI[2] = "C:\HOUTRANS\BASE02\TIME&DIST.MAT"
;
FILEI ZDATI[2] = "C:\HOUTRANS\BASE02\INPUT_DATA\PR_INTRA_02.DBF"
FILEI ZDATI[1] = "C:\HOUTRANS\BASE02\BTGEN00A.DBF"
FILEO MATO[1] = "C:\HOUTRANS\BASE02\BDMAT00A.MAT",
MO=11 NAME=SYN_Internal
;
PARAMETERS ZONES=270
DEC=2
;
MW[11]=0.0
;
Alpha=0.584
Beta =0.598
;
Gama =0.728
Kfac =0.0027
;
JLOOP
;
GEN = P1[I]
ATT = A1[J]
;
IMPED = MI.2.2 - INTRA_DIST[I] - INTRA_DIST[J]
;
OBS_INDX = MI.1.8
;
MW[11][J]=0.0
IF(I<>J && IMPED>0.0 && OBS_INDX=1) ; Skip Intrazonal AND Unobserved Cells
;
MW[11] = KfAC * ((GEN^Alpha) * (ATT^Beta) / (IMPED^Gama) )
;
ENDIF
;
ENDJLOOP

IF (I<266) MW[11][I]=(P1[I]+A1[I])*.5*PR_INT_02 ; Infill Intras for Internal

Zones

ENDRUN

A2-10
 The Study on Urban Transport Master Plan and Feasibility Study in HCM Metropolitan Area (HOUTRANS)
Final Report
VOLUME 4: Transport Database

2. Distribute Internal <> External and External<>External Trips – Script File

; Do not change filenames or add or remove FILEI/FILEO statements using an editor.
Use Cube/Application Manager.
RUN PGM=MATRIX PRNFILE="C:\HOUTRANS\BASE02\BDMAT00B.PRN" MSG='Base Year Dist
Add-(Ext x Ext) Trips'
FILEI ZDATI[1] = "C:\HOUTRANS\BASE02\BTGEN00A.DBF"
;
FILEI MATI[1] = "C:\HOUTRANS\BASE02\INPUT_DATA\OBS_PRSN_TRIPS_02.MAT"
;
FILEI MATI[2] = "C:\HOUTRANS\BASE02\BDMAT00A.MAT"
;
FILEO MATO[1] = "C:\HOUTRANS\BASE02\BDMAT00B.MAT",
MO=11 NAME=SYN_Temp
;
;
PARAMETERS ZONES=270
DEC=2
;
;
MW[1]=MI.1.7
MW[2]=MI.2
MW[11]=MW[2]
;
JLOOP J=266,270
MW[11][J] = MW[1][J] * P2[I]
ENDJLOOP
;
JLOOP
IF(I>265) MW[11][J] = MW[1][J]* P2[I]
ENDJLOOP
;
;
ENDRUN

A2-11
 The Study on Urban Transport Master Plan and Feasibility Study in HCM Metropolitan Area (HOUTRANS)
Final Report
VOLUME 4: Transport Database

3. Control Total Number of Distributed Trips – Script File

; Do not change filenames or add or remove FILEI/FILEO statements using an editor.
Use Cube/Application Manager.
RUN PGM=MATRIX PRNFILE="C:\HOUTRANS\BASE02\BDMAT00C.PRN" MSG='Base Year Dist
(IntxInt) Trips'
;
FILEI MATI[1] = "C:\HOUTRANS\BASE02\INPUT_DATA\OBS_PRSN_TRIPS_02.MAT"
;
FILEI MATI[2] = "C:\HOUTRANS\BASE02\BDMAT00B.MAT"
;
FILEI ZDATI[1] = "C:\HOUTRANS\BASE02\BTGEN00A.DBF"
;
FILEO MATO[1] = "C:\HOUTRANS\BASE02\BDMAT00C.MAT",
MO=11 NAME=SYN_All_Modes
;
MW[1]=MI.1.7
MW[2]=MI.2
;
MW[11]=0.0
;
OBS_TOT=0.0
TMP_TOT=0.0
;
LOOP II=1,ZONES
LOOP JJ=1,ZONES
OBS_TOT = OBS_TOT+MATVAL(1,7,II,JJ)
TMP_TOT = TMP_TOT+MATVAL(2,1,II,JJ)
ENDLOOP
ENDLOOP
;
ADJ_FACT = OBS_TOT * P2[270] / TMP_TOT
;
JLOOP
MW[11][J] = MW[2][J] * ADJ_FACT
ENDJLOOP
ENDRUN

A2-12
 The Study on Urban Transport Master Plan and Feasibility Study in HCM Metropolitan Area (HOUTRANS)
Final Report
VOLUME 4: Transport Database

2.5 MODAL SPLIT MODELS

The Modal Split model divides the total Trip Distributed matrix into six modes: 1-Wlak, 2-Bicycle,
3-Motorcycle, 4-Car, 5-Truck and 6-Bus in the case of base year model run. For forecast years the
there are no Person Trips by Bicycle or Truck. As it was assumed that person trips by these modes
will be negligible. The Model is executed by a single run of Program MATRIX, for both walk and
non-walk (Vehicle) Trip modal Split models.

Program Used: MATRIX

Input Files:

1. Base Year Observed Trips Matrix and Observed Cells Index File;
2. Time & Distance Skim File; and

3. Full Distributed Total trip Matrix.

Output File:

1. Output Trip Person Trip Matrices by Mode and Total PCU matrix.

A2-13
 The Study on Urban Transport Master Plan and Feasibility Study in HCM Metropolitan Area (HOUTRANS)
Final Report
VOLUME 4: Transport Database

Script File for Modal Split Models:

; Do not change filenames or add or remove FILEI/FILEO statements using an editor.

Use Cube/Application Manager.
RUN PGM=MATRIX PRNFILE="C:\HOUTRANS\BASE02\BMMAT00A.PRN" MSG='Modal Split
Model'
;
FILEI ZDATI[1] = "C:\HOUTRANS\BASE02\INPUT_DATA\INTRA_DIST.DBF"
;
FILEI MATI[1] = "C:\HOUTRANS\BASE02\INPUT_DATA\OBS_PRSN_TRIPS_02.MAT"
FILEI MATI[2] = "C:\HOUTRANS\BASE02\TIME&DIST.MAT"
;
FILEI MATI[3] = "C:\HOUTRANS\BASE02\BDMAT00C.MAT"
;
FILEO MATO[1] = "C:\HOUTRANS\BASE02\BMMAT00A.MAT",
MO=31-38 NAME=WALK, BCY, MCY, CAR, TRUCK, BUS, TOT_PRSN, TOT_PCU
;
;
; Input Trip Matrices
;
MW[1]=MI.1.1
MW[2]=MI.1.2
MW[3]=MI.1.3
MW[4]=MI.1.4
MW[5]=MI.1.5
MW[6]=MI.1.6
MW[7]=MI.1.7
MW[8]=MI.1.8
MW[9]=0.0
; Modal Split Prob Mxs
MW[21]=0.0 ; PR_WLK
MW[22]=0.0 ; PR_BCY
MW[23]=0.0 ; PR_MCY
MW[24]=0.0 ; PR_Car
MW[25]=0.0 ; PR_Truck
MW[26]=0.0 ; PR_Bus
MW[27]=0.0 ; PR_ALL
;
; OUTPUT TRIP Matrices
;
;
MW[31]=0.0 ; Output WALK
MW[32]=0.0 ; Output BCY
MW[33]=0.0 ; Output MCY
MW[34]=0.0 ; Output Car
MW[35]=0.0 ; Output Truck
MW[36]=0.0 ; Output Bus
MW[37]=0.0 ; Tot_PRSN Output
MW[38]=0.0 ; Tot_PCU
;
; Input Trip Matrices Set Intras=0 as not required any more
;
MW[1][I]=0.0
MW[2][I]=0.0
MW[3][I]=0.0
MW[4][I]=0.0
MW[5][I]=0.0
MW[6][I]=0.0
MW[7][I]=0.0
;

A2-14
 The Study on Urban Transport Master Plan and Feasibility Study in HCM Metropolitan Area (HOUTRANS)
Final Report
VOLUME 4: Transport Database

MW[41]=MI.3.1
MW[41][I]=0.0
;
MW[9]=0.0 ; Synth Total Less Walk Trips for MS_Model
;
; COMPUTE WALK TRIPS AS OBS PROP & CALC BALANCE TRIPS
;
JLOOP
IF (MW[7][J]>0.0) MW[21][J]=MW[1][J]/MW[7][J]
MW[31][J] = MW[21][J] * MW[41][J]
MW[42][J] = MW[41][J] - MW[31][J]
ENDJLOOP
;
;
;
; COMPUTE I<>J TRIPS PROP.
;
II_Dist = INTRA_DIST[I]
II_Time = II_DIST/5.0 ; Access Walk Speed=5.0 Time in Hours @ ORIGIN
;
;
JLOOP
;
IF (I<>J) ; FOR INTER-ZONAL TRIPS ONLY
;
;
JJ_Dist = INTRA_DIST[J]
JJ_Time = JJ_DIST/5.0 ; Access Walk Speed=5.0 Time in Hours @ DEST
;
;
IJTIME = TIME[J]/60.0 - II_TIME - JJ_TIME ; Time in Hours
IJDIST = DIST[J] - II_DIST - JJ_DIST
;
; GC SKIMS & CALCULATE PROP TRIPS

;
MW[22][J] = IJTIME*0.464 + (( IJDIST*0.0034+0.039)/1.14) ; GC_BCY
MW[22][J] = 1 / (1 + (EXP( 0.598*MW[22][J] + 1.814)) )
;
MW[23][J] = IJTIME*0.823 + (( IJDIST*0.0268+0.083)/1.51) ; GC_MCY
MW[23][J] = 1 / (1 + (EXP( 0.829*MW[23][J] - 1.907)) )
;
MW[24][J] = IJTIME*1.732 + (( IJDIST*0.0705+0.314)/1.96) ; GC_CAR
MW[24][J] = 1 / (1 + (EXP( 0.139*MW[24][J] + 2.950)) )
;
MW[25][J] = IJTIME*1.025 + (( IJDIST*0.1535+0.350)/1.98) ; GC_TRUCK
MW[25][J] = 1 / (1 + (EXP( 0.133*MW[25][J] + 3.963)) )
;
IF (IJDIST <= 4.0) MW[26][J] = IJTIME*0.525 + 0.067 ; BUS TRIP <=4.0 KM
IF (IJDIST > 4.0) MW[26][J] = IJTIME*0.525 + 0.067 + (IJDIST-4.0)*.016 ;
BUS TRIP > 4.0 KM
MW[26][J] = 1 / (1 + (EXP( 0.513*MW[26][J] + 3.052)) )
;
ENDIF
;
ENDJLOOP ; END OF TRIP PROP CALCULATIONS
;
; Calculte Sum_Prop
JLOOP
MW[27] = MW[22]+MW[23]+MW[24]+MW[25]+MW[26]
ENDJLOOP

A2-15
 The Study on Urban Transport Master Plan and Feasibility Study in HCM Metropolitan Area (HOUTRANS)
Final Report
VOLUME 4: Transport Database

;
; Calculte Prop where Sum_prop>1.0
JLOOP
IF (MW[27][J]>0.0))
MW[22][J] = MW[22][J] / MW[27][J]
MW[23][J] = MW[23][J] / MW[27][J]
MW[24][J] = MW[24][J] / MW[27][J]
MW[25][J] = MW[25][J] / MW[27][J]
MW[26][J] = MW[26][J] / MW[27][J]
ENDIF
ENDJLOOP
;
; COMPUTE OTHER TRIPS AS PROP OF BALANCE TRIPS
;
JLOOP
MW[32][J] = MW[22][J] * MW[42][J]
MW[33][J] = MW[23][J] * MW[42][J]
MW[34][J] = MW[24][J] * MW[42][J]
MW[35][J] = MW[25][J] * MW[42][J]
MW[36][J] = MW[26][J] * MW[42][J]
ENDJLOOP
;
MW[37] = MW[31]+MW[32]+MW[33]+MW[34]+MW[35]+MW[36]
;
;
MW[38] = (MW[32] / 1.14 * 0.2) +
(MW[33] / 1.51 * 0.2) +
(MW[34] / 1.96 * 1.0) +
(MW[35] / 1.98 * 2.5) +
(MW[36] / 8.18 * 2.5)
;
;
ENDRUN

A2-16
 The Study on Urban Transport Master Plan and Feasibility Study in HCM Metropolitan Area (HOUTRANS)
Final Report
VOLUME 4: Transport Database

2.6 TRAFFIC ASSIGNMENT AND ANALYSIS

The Traffic Assignment Model Assigns the total PCU trip matrix to the highway network, using
capacity restraint procedure. Figure below shows 4-assignments. The first assignment model is for
24-hour trips, and it uses incremental assignment method and assigns the total traffic in 10
iterations in 10 equal increments. The output Loaded network is then processed to compute
various network statistics. It is up to the user, what statistics he/she wishes to produce. The script
file can be developed according to the outputs required.

The next three sets of assignments are for three one hourly assignments, representing average
AM, PM and Off peak hour traffic volumes. The peak matrices for base year are derived using
peak hour factors as described in the 3rd Stage model documentation in Section 3.6. For these
assignments same methodology was adopted. In each case equilibrium assignment techniques
was used, and each vehicle class was assigned separately to time paths, so that each vehicle
results could be analysed individually.

Step-1: Program Used: Highway for Traffic Assignment

Input Files:

1. Modal Split Model Output Total PCU matrix; and

2. Loaded Network File

Output Files:
1. Time & Distance Skims (Optional); and

2. Output Loaded Network File

A2-17
 The Study on Urban Transport Master Plan and Feasibility Study in HCM Metropolitan Area (HOUTRANS)
Final Report
VOLUME 4: Transport Database

Step-1 Script File for HIGHWAY Program:

; Do not change filenames or add or remove FILEI/FILEO statements using an editor.
Use Cube/Application Manager.
RUN PGM=HIGHWAY PRNFILE="C:\HOUTRANS\BASE02\BAHWY00A.PRN" MSG='Assign Synth
Mx'
FILEI MATI[1] = "C:\HOUTRANS\BASE02\BMMAT00A.MAT"
FILEI NETI = "C:\HOUTRANS\BASE02\BNNET00A.NET"
;
FILEO NETO = "C:\HOUTRANS\BASE02\BAHWY00A.NET",
EXCLUDE=T_
FILEO MATO[1] = "C:\HOUTRANS\BASE02\BAHWY00A.MAT",
MO=1-2 Name=TIME_OUT, DIST_OUT
; ----- SET Parameters

PARAMETERS COMBINE=SUM, FRACTIONS = 10 * 0.10 ; Assignment 10 Iters Load

10% Each Time
;
; PARAMETERS COMBINE=equ, MAXITERS=10, GAP=0.001 ; Assignment EQU, 10 Iters
Load MAX
;
Time_Par=1.0 ; Time used
Dist_Par=0.0 ; Note Dist_Par=0.0
;
; ----- SET CAPACITY and group links
;
PROCESS PHASE=LINKREAD
;
C = LI.CAPACITY * 1.0
;
IF (li.CLASS = 1-30) LINKCLASS = 1 ; set link classes for roads
IF (li.CLASS = 31) LINKCLASS = 2 ; set link classes for CC
;
Min_Speed=Li.Speed*0.1
;
ENDPROCESS
;
;
PROCESS PHASE=ILOOP ; Build, Skim & Load All
Zones I=1,270
;
MW[11]=MI.1.8
;
PATHLOAD PATH=COST, ; Build Cost paths
;
MW[1]=PATHTRACE(TIME), MW[2]=PATHTRACE(LI.DISTANCE), ; Build Time & Dist Skims
;
VOL[1]=MI.1.8 ; Assign PCU Matrix
;
; Store Each Iteration Assignment to Vol Fields 11-20 for Iterations=1-10
;
IF(ITERATION=1) PATHLOAD PATH=COST, VOL[11]=MW[11]
IF(ITERATION=2) PATHLOAD PATH=COST, VOL[12]=MW[11]
IF(ITERATION=3) PATHLOAD PATH=COST, VOL[13]=MW[11]
IF(ITERATION=4) PATHLOAD PATH=COST, VOL[14]=MW[11]
IF(ITERATION=5) PATHLOAD PATH=COST, VOL[15]=MW[11]
IF(ITERATION=6) PATHLOAD PATH=COST, VOL[16]=MW[11]
IF(ITERATION=7) PATHLOAD PATH=COST, VOL[17]=MW[11]
IF(ITERATION=8) PATHLOAD PATH=COST, VOL[18]=MW[11]
IF(ITERATION=9) PATHLOAD PATH=COST, VOL[19]=MW[11]
IF(ITERATION=10) PATHLOAD PATH=COST, VOL[20]=MW[11]

A2-18
 The Study on Urban Transport Master Plan and Feasibility Study in HCM Metropolitan Area (HOUTRANS)
Final Report
VOLUME 4: Transport Database

;
;
MW[1][I]=0 ; Set Time Skim Intra=0.0
MW[2][I]=0 ; Set Dist Skim Intra=0.0
;
;
ENDPHASE ; End of Loop
;
;
Process PHASE = ADJUST ; Adjust Time from S/F Relationship Called for each
Link Before Path Build
;
Function {
;
tc[1]=(Li.distance/( max((Speed-((Speed-Min_Speed)*v/c)),Min_Speed)))*60
;
tc[2]=(Li.distance/( max((Speed-((Speed-Speed*1.0)*v/c)),Speed*1.0)))*60
;
V=VOL[1] ; Define total as Vol 1 only.
;
cost[1] = Time*Time_Par + Li.Distance*Dist_Par
cost[2] = Time*Time_Par + Li.Distance*Dist_Par }
;
ENDPHASE
;
;
ENDRUN

A2-19
 The Study on Urban Transport Master Plan and Feasibility Study in HCM Metropolitan Area (HOUTRANS)
Final Report
VOLUME 4: Transport Database

Step-2: Program Used: NETWORK for the Analyses of the Assigned Traffic

Input Files:
3. Loaded Network File output by the HIGHWAY Program at Step-1

Output Files:

1. Loaded Network Binary File with Additional Attributes; and

2. Output Loaded Network “DBF ”File for Analyses using XL Spreadsheet or similar
software.

Step-2 Script File for NETWORK Program:

; Do not change filenames or add or remove FILEI/FILEO statements using an editor.
Use Cube/Application Manager.
RUN PGM=NETWORK PRNFILE="C:\HOUTRANS\BASE02\BANET00A.PRN" MSG='Compute Stats
& Output Network'
FILEO NETO = "C:\HOUTRANS\BASE02\BANET00A.NET",
EXCLUDE=T_ ; OUTPUT Network
;
FILEO LINKO = "C:\HOUTRANS\BASE02\OUTPUT_LNK.DBF"
FILEI LINKI[1] = "C:\HOUTRANS\BASE02\BAHWY00A.NET"
;
SPEED_1=INT(DISTANCE/(TIME_1/60)+.5); Output Network Speed - relative to
Time_1
;
TIME_1=(INT(TIME_1*100+.5))*.01 ; Round Final Output Time (mins)
;
PCU_KM = INT(V1_1*DISTANCE*10+.5)*.1 ; Network Travelled PCU*Kms
PCU_HR = INT(V1_1*TIME_1/.6+.5)*.01 ; Network Travelled PCU*Final Travel Time
;
;
V_1 =(INT(V_1*10.0+.5))*.1 ; Round Final Output Volume PCUs
;
VC_1=(INT(VC_1*1000+.5))*.001; Round Final Output V/C Ratio, 1-Way
;
V1_1 =(INT(V1_1*.01+.5))*100; Round Final Output 1-Way Volume PCUs in Unit of
100 for Plotting
;
VT_1 =(INT(VT_1*.01+.5))*100; Round Final Output 2-Way Volume PCUs in Unit of
100 for Plotting
;
; Calculate Cumulative Vols for 10_iters.
;
V11_1=V11_1
V12_1=V12_1+V11_1
V13_1=V13_1+V12_1
V14_1=V14_1+V13_1
V15_1=V15_1+V14_1
V16_1=V16_1+V15_1
V17_1=V17_1+V16_1
V18_1=V18_1+V17_1
V19_1=V19_1+V18_1
V20_1=V20_1+V19_1
;
;
; Calculate Cumulative V/C for 10_iters.
;

A2-20
 The Study on Urban Transport Master Plan and Feasibility Study in HCM Metropolitan Area (HOUTRANS)
Final Report
VOLUME 4: Transport Database

IF(V11_1>0.0) VC_11 = V11_1 / CAPACITY

IF(V12_1>0.0) VC_12 = V12_1 / CAPACITY
IF(V13_1>0.0) VC_13 = V13_1 / CAPACITY
IF(V14_1>0.0) VC_14 = V14_1 / CAPACITY
IF(V15_1>0.0) VC_15 = V15_1 / CAPACITY
IF(V16_1>0.0) VC_16 = V16_1 / CAPACITY
IF(V17_1>0.0) VC_17 = V17_1 / CAPACITY
IF(V18_1>0.0) VC_18 = V18_1 / CAPACITY
IF(V19_1>0.0) VC_19 = V19_1 / CAPACITY
IF(V20_1>0.0) VC_20 = V20_1 / CAPACITY
;
;
Min_Speed=Speed*0.1
IF(CAPINX=2) MIN_SPEED=SPEED
;
; Calculate Speed at V/C for 10_iters.
;
SPEED_11 =(( max((Speed-((Speed-Min_Speed)*VC_11)),Min_Speed)))
SPEED_12 =(( max((Speed-((Speed-Min_Speed)*VC_12)),Min_Speed)))
SPEED_13 =(( max((Speed-((Speed-Min_Speed)*VC_13)),Min_Speed)))
SPEED_14 =(( max((Speed-((Speed-Min_Speed)*VC_14)),Min_Speed)))
SPEED_15 =(( max((Speed-((Speed-Min_Speed)*VC_15)),Min_Speed)))
SPEED_16 =(( max((Speed-((Speed-Min_Speed)*VC_16)),Min_Speed)))
SPEED_17 =(( max((Speed-((Speed-Min_Speed)*VC_17)),Min_Speed)))
SPEED_18 =(( max((Speed-((Speed-Min_Speed)*VC_18)),Min_Speed)))
SPEED_19 =(( max((Speed-((Speed-Min_Speed)*VC_19)),Min_Speed)))
SPEED_20 =(( max((Speed-((Speed-Min_Speed)*VC_20)),Min_Speed)))
;
; Calculate Average Speed for 10_iters.
;

SPEED_AV=(SPEED_11+SPEED_12+SPEED_13+SPEED_14+SPEED_15+
SPEED_16+SPEED_17+SPEED_18+SPEED_19+SPEED_20)*.1
SPEED_AV=INT(SPEED_AV*10+0.5)*.1
;
; Calculate Average Time for 10_iters.
;
IF (SPEED_AV>0.0) TIME_AV=INT(DISTANCE/SPEED_AV*6000+.5)*.01
;
; Calculate PCU*Hrs for Average Time for 10_iters.
;
AV_PCU_HR = INT(V_1*TIME_AV/.6+.5)*.01
;
; Calculate PCU*Hrs for Excluding Connectors for Av of 10_iters.
;
LNK_PCU_HR=AV_PCU_HR
IF(CAPINX=2) LNK_PCU_HR=0.0
;
; Calculate PCU*kms for Excluding Connectors.
;
LNK_PCU_KM=PCU_KM
IF(CAPINX=2) LNK_PCU_KM=0.0;
;
;
;
ENDRUN

A2-21
 The Study on Urban Transport Master Plan and Feasibility Study in HCM Metropolitan Area (HOUTRANS)
Final Report
VOLUME 4: Transport Database

ANNEX 3 CUBE/VOYAGER MODEL DESCRIPTION (2)

3.1 TRIP GENERATION AND ATTRACTION MODELS

The CUBE/Voyager Model is shown below. The model shown is for the year 2020. There are five
model runs, one for each trip purposes. Each program run is similar in action and has the same
Input data file. The Outputs are also similar, except that each is for a different trip purpose. The
model is run only once for each trip purpose.

Step-1: Program Used: GENERATIONS for Tripend Claculations

Input Files:

1. 20 20 Planning Data File created in DBF format using XL_Spreadsheet.

This File must contain the following data for each of the 1-265 internal zones of the study area. All
variable names as setout in the DBF file should remain the same and are explained below
1. Z: Zone Number
2. LU: Code for the zone as given in Table 1.2 in Annex 1.
3. AREA: Zone Area in Hectare
4. POP: Total Population in the Zone
5. HHTOT: Total Number of Households in the Zone
6. CARHH: Number of Car Owning Households in the Zone
7. MC2HH: 2+ Motorcycle Owning Households in the Zone
8. MC1HH: 1 Motorcycle Owning Households in the Zone
9. NOVHH: Number of Households Owning No vehicles

A3-1
 The Study on Urban Transport Master Plan and Feasibility Study in HCM Metropolitan Area (HOUTRANS)
Final Report
VOLUME 4: Transport Database

10. RPW: Number of Residents in the Zone Employed in Primary Sector Jobs
11. RSW: Number of Residents in the Zone Employed in Secondary Sector Jobs
12. RTW: Number of Residents in the Zone Employed in Tertiary Sector Jobs
13. RPS: Number of Resident Students in the Zone Attending Primary Education
14. RSS: Number of Resident Students in the Zone Attending Secondary Education
15. RTS: Number of Resident Students in the Zone Attending Tertiary Education
16. JPW: Number of Primary Sector Jobs in the Zone
17. JSW: Number of Secondary Sector Jobs in the Zone
18. JTW: Number of Tertiary Sector Jobs in the Zone
19. SPS: Number of Primary Education Places in the Zone
20. SSS: Number of Secondary Education Places in the Zone
21. STS: Number of Tertiary Education Places in the Zone
22. PCAR: Proportion of Population in Car Owing HH which have a vehicle Available for Travel
23. P2MC: Proportion of Population in 2+M/C Owing HH which have a vehicle Available for
Travel
24. P1MC: Proportion of Population in 1 M/C Owing HH which have a vehicle Available for
Travel
Notes
Same data file is input to all five model runs. All totals must be internally consistent. Program
makes no checks for the totals.

Output Files:
1. Program produces a Single DBF file containing seven sets of trip Generations and
Attractions as defined below. All trip attractions are balanced to trip generations for the
same purpose and HH category. The over all trip generations are controlled to study
area-wide trip control total for the that purpose and the overall study area trip control totals.
The control totals. The individual trip generations controls are calculated automatically,
however study area trip control total is calculated exogenously and entered in the Script
file. The Seven sets of trip generations and attractions are:

P1: Trip Generations for Car Owning Households who had a Vehicle Available
A1: Trip Attractions for Car Owning Households who had a Vehicle Available

P2: Trip Generations for Car Owning Households who had NO Vehicle Available
A2: Trip Attractions for Car Owning Households who had NO Vehicle Available

P3: Trip Generations for 2+M/C Owning Households who had a Vehicle Available
A3: Trip Attractions for 2+M/C Owning Households who had a Vehicle Available

P4: Trip Generations for 2+M/C Owning Households who had NO Vehicle Available
A4: Trip Attractions for 2+M/C Owning Households who had NO Vehicle Available
P5: Trip Generations for 1M/C Owning Households who had a Vehicle Available
A5: Trip Attractions for 1M/C Owning Households who had a Vehicle Available

A3-2
 The Study on Urban Transport Master Plan and Feasibility Study in HCM Metropolitan Area (HOUTRANS)
Final Report
VOLUME 4: Transport Database

P6: Trip Generations for 1M/C Owning Households who had NO Vehicle Available
A6: Trip Attractions for 1M/C Owning Households who had NO Vehicle Available

P7: Trip Generations for NO Vehicle NO Owning Households

A7: Trip Attractions for NO Vehicle Owning Households
The values for zones 266-270 will be set to Zero for all sets.

Step-1 Script File for GENERATION Program:

; Do not change filenames or add or remove FILEI/FILEO statements using an editor.
Use Cube/Application Manager.
RUN PGM=GENERATION PRNFILE="C:\HOUTRANS\MODEL_S3\WRK_TEND.PRN" MSG='Calc
To_Work TE'
;
FILEI ZDATI[1] = "C:\HOUTRANS\MODEL_S3\INPUTDATA\PD20_T4R.DBF"
;
FILEO PAO[1] = "C:\HOUTRANS\MODEL_S3\WRK_TEND.DAT", DBF=T, LIST=Z(4.0),
P[1](7.0), A[1](7.0), P[2](7.0), A[2](7.0), P[3](7.0), A[3](7.0),
P[4](7.0), A[4](7.0), P[5](7.0), A[5](7.0), P[6](7.0), A[6](7.0), P[7](7.0),
A[7](7.0)
;
;
PARAMETERS ZONES=270, MAXPURPS=7
;
;
;

PROCESS PHASE=ILOOP ; Calculate Trip_Ends

& Factors
;
; Calc Total Population by Vehicle Ownership Category to Get Trip Control Totals.
;
IF (Z>265) GOTO EXT1
;

Pop_V1 = Pop * (CarHH/HHTot)*Pcar

Pop_V2 = Pop * (CarHH/HHTot)*(1-Pcar)
;
Pop_V3 = Pop * (MC2HH/HHTot)*P2MC
Pop_V4 = Pop * (MC2HH/HHTot)*(1-P2MC)
;
Pop_V5 = Pop * (MC1HH/HHTot)*P1MC
Pop_V6 = Pop * (MC1HH/HHTot)*(1-P1MC)
;
Pop_V7 = Pop -Pop_V1-Pop_V2-Pop_V3-Pop_V4-Pop_V5-Pop_V6
;
IF (Pop_V7<0) Pop_V7=0
;
; Calc Total Pop by Veh_Ownership
;
TP_V1 = TP_V1 + Pop_V1
TP_V2 = TP_V2 + Pop_V2
TP_V3 = TP_V3 + Pop_V3
TP_V4 = TP_V4 + Pop_V4
TP_V5 = TP_V5 + Pop_V5
TP_V6 = TP_V6 + Pop_V6
TP_V7 = TP_V7 + Pop_V7
;

A3-3
 The Study on Urban Transport Master Plan and Feasibility Study in HCM Metropolitan Area (HOUTRANS)
Final Report
VOLUME 4: Transport Database

; Calc Zonal Trip Prods.

;
Tot_RW = RPW + RSW + RTW ; Calc Total Res_Workers
;
Pxxx = 33.1 + 1.392 * Tot_RW
P[1] = Pxxx * Pop_V1 /(Pop_V1+POP_V2)
P[2] = Pxxx - P[1]
;
;
Pxxx = 31.9 + 1.306 * RSW + 1.179 * RTW
P[3] = Pxxx * Pop_V3 /(Pop_V3+POP_V4)
P[4] = Pxxx - P[3]
;
Pxxx = 28.6 + 1.292 * RPW + 1.150 * RSW + 0.875 * RTW
P[5] = Pxxx * Pop_V5 /(Pop_V5+POP_V6)
P[6] = Pxxx - P[5]
;
P[7] = 1.054 * Tot_RW
;
; Calc Zonal Trip Atts.
;
A[1] = 53.7 + 0.174 * JPW + 2.682 * JSW + 1.057 * JTW
A[2] = A[1]
;
A[3] = 39.5 + 1.356 * JSW + 1.276 * JTW
A[4] = A[3]
;
A[5] = 1.228 * JPW + 1.202 * JSW + 0.951 * JTW
A[6] = A[5]
;
A[7] = 6.3 + 1.453 * JPW + 0.995 * JSW + 0.791 * JTW
;
; Calculate Total Prods.
;
TT_P1 = TT_P1 + P[1]
TT_P2 = TT_P2 + P[2]
TT_P3 = TT_P3 + P[3]
TT_P4 = TT_P4 + P[4]
TT_P5 = TT_P5 + P[5]
TT_P6 = TT_P6 + P[6]
TT_P7 = TT_P7 + P[7]
;
;
:EXT1
;
ENDPHASE
;
PROCESS PHASE=ADJUST
;
;
; Calc Prod Control Totals FOR TO-WORK TRIPS
;
TRP_RATE_V1 = 0.8792
TRP_RATE_V2 = 0.1899
TRP_RATE_V3 = 0.7574
TRP_RATE_V4 = 0.1800
TRP_RATE_V5 = 0.6392
TRP_RATE_V6 = 0.2114
TRP_RATE_V7 = 0.4202
;
CTL_FACT = 0.943634

A3-4
 The Study on Urban Transport Master Plan and Feasibility Study in HCM Metropolitan Area (HOUTRANS)
Final Report
VOLUME 4: Transport Database

;
CT_P1 = TP_V1 * TRP_RATE_V1 * CTL_FACT
CT_P2 = TP_V2 * TRP_RATE_V2 * CTL_FACT
CT_P3 = TP_V3 * TRP_RATE_V3 * CTL_FACT
CT_P4 = TP_V4 * TRP_RATE_V4 * CTL_FACT
CT_P5 = TP_V5 * TRP_RATE_V5 * CTL_FACT
CT_P6 = TP_V6 * TRP_RATE_V6 * CTL_FACT
CT_P7 = TP_V7 * TRP_RATE_V7 * CTL_FACT
;
JLOOP ; Control Zonal Prods to Control Totals
;
IF (J<266) P[1] = CT_P1 / TT_P1 * P[1]
IF (J<266) P[2] = CT_P2 / TT_P2 * P[2]
IF (J<266) P[3] = CT_P3 / TT_P3 * P[3]
IF (J<266) P[4] = CT_P4 / TT_P4 * P[4]
IF (J<266) P[5] = CT_P5 / TT_P5 * P[5]
IF (J<266) P[6] = CT_P6 / TT_P6 * P[6]
IF (J<266) P[7] = CT_P7 / TT_P7 * P[7]
;
ENDJLOOP
;
; Set External Trip_Ends=0
;
P[1][266]=0 P[1][267]=0 P[1][268]=0 P[1][269]=0 P[1][270]=0
P[2][266]=0 P[2][267]=0 P[2][268]=0 P[2][269]=0 P[2][270]=0
P[3][266]=0 P[3][267]=0 P[3][268]=0 P[3][269]=0 P[3][270]=0
P[4][266]=0 P[4][267]=0 P[4][268]=0 P[4][269]=0 P[4][270]=0
P[5][266]=0 P[5][267]=0 P[5][268]=0 P[5][269]=0 P[5][270]=0
P[6][266]=0 P[6][267]=0 P[6][268]=0 P[6][269]=0 P[6][270]=0
P[7][266]=0 P[7][267]=0 P[7][268]=0 P[7][269]=0 P[7][270]=0
;
A[1][266]=0 A[1][267]=0 A[1][268]=0 A[1][269]=0 A[1][270]=0
A[2][266]=0 A[2][267]=0 A[2][268]=0 A[2][269]=0 A[2][270]=0
A[3][266]=0 A[3][267]=0 A[3][268]=0 A[3][269]=0 A[3][270]=0
A[4][266]=0 A[4][267]=0 A[4][268]=0 A[4][269]=0 A[4][270]=0
A[5][266]=0 A[5][267]=0 A[5][268]=0 A[5][269]=0 A[5][270]=0
A[6][266]=0 A[6][267]=0 A[6][268]=0 A[6][269]=0 A[6][270]=0
A[7][266]=0 A[7][267]=0 A[7][268]=0 A[7][269]=0 A[7][270]=0
;

;
BALANCE A2P=1-7 ; Set Att_Total to Pro_Totals ALL
7-SetS.
;
;
ENDPHASE ; End of Adjustments
;
;
;
ENDRUN

Step-2-5 Script Files for GENERATION Program:

These files for other four trip purposes are not included here as these files are similar and only the
parameter values change.

A3-5
 The Study on Urban Transport Master Plan and Feasibility Study in HCM Metropolitan Area (HOUTRANS)
Final Report
VOLUME 4: Transport Database

3.2 TRIP DISTRIBUTION MODELS

Build Time & Distance Skims Matrices
The key input to the distribution models is the zone-to-zone travel distances. These are derived
from the network model as shown below. The complete methodology has been described in
Annex 2.1 above. The only difference is that the Input network file represents the future year
network under test. The remaining process is the same as shown below and is not repeated here.

Trip Distribution Models

Figure below illustrates the trip distribution models for five trip purposes. Each trip purpose trip
distribution is performed in two steps, which are described next:

Step-1: Initial Trip Distribution using Program MATRIX:

In this process trips are internal<>internal trips for Tij for interzonal trips are estimated as given in
equation in Section 3.3. In a single run of the program trips are estimated for one trip purpose but
for all seven household categories in a single run.

Program Used: MATRIX

A3-6
 The Study on Urban Transport Master Plan and Feasibility Study in HCM Metropolitan Area (HOUTRANS)
Final Report
VOLUME 4: Transport Database

Input Files:

1. Distance Skim File;

2. Trip Generations and Attractions as P1 to P& & A1 to A7 for each of the Seven HH
categories output by the Tripend Model.

Output File:

1. Out Internal <> Internal Trip Matrix File Containing seven trip tables , one for each
household categories for a single trip purpose. Excluding Intrazonal Trips

Step-1 Script File for MATRIX Program:

; Do not change filenames or add or remove FILEI/FILEO statements using an editor.
Use Cube/Application Manager.
RUN PGM=MATRIX PRNFILE="C:\HOUTRANS\MODEL_S3\HDMAT00A.PRN"
MSG='Work_Trip_Syn'
FILEI MATI[1] = "C:\HOUTRANS\MODEL_S3\TIME&DIST.MAT"
;
FILEI ZDATI[1] = "C:\HOUTRANS\MODEL_S3\WRK_TEND.DAT"
;
;
FILEO MATO[1] = "C:\HOUTRANS\MODEL_S3\HDMAT00A.MAT",
Dec=2,
MO=11-17 NAME=WCAROW,WCARNO, W2MCOW,W2MCNO, W1MCOW,W1MCNO, WNOVEH
;
PARAMETERS ZONES=270
DEC=2
;
MW[11]=0.0 MW[12]=0.0 MW[13]=0.0 MW[14]=0.0 MW[15]=0.0 MW[16]=0.0, MW[17]=0.0
;
; Set Distribution Model Parameters
;
K1 = 1.390 ; Car Owning
AL1 = 0.332
BT1 = 0.309
GM1 = -.0179
;
K2 = 3.020 ; Car Owning -Non_User
AL2 = 0.329
BT2 = 0.316
GM2 = -.0261
;
;
K3 = 0.732 ; 2_MC Owning
AL3 = 0.326
BT3 = 0.266
GM3 = -.2900
;
K4 = 0.634 ; 2_MC Owning -Non_User
AL4 = 0.283
BT4 = 0.369
GM4 = -.0289
;
;
K5 = 0.434 ; 1_MC Owning
AL5 = 0.407
BT5 = 0.297
GM5 = -.1440

A3-7
 The Study on Urban Transport Master Plan and Feasibility Study in HCM Metropolitan Area (HOUTRANS)
Final Report
VOLUME 4: Transport Database

;
K6 = 0.382 ; 1_MC Owning -Non_User
AL6 = 0.377
BT6 = 0.357
GM6 = -.0481
;
;
K7 = 1.210 ; NO_Veh Owning
AL7 = 0.298
BT7 = 0.339
GM7 = -.0812
;
; Distribute Trips for All Seven Categories of HH's
;
JLOOP
;
GEN = P1[I]
ATT = A1[J]
IMPED = MI.1.2[J]
MW[11][J]=0.0
IF(I<>J && IMPED>0.0 ) ; Skip Intrazonal
MW[11][J] = K1 * ((GEN^AL1) * (ATT^BT1) /(IMPED^(-GM1) ) )
ENDIF
;
GEN = P2[I]
ATT = A2[J]
IMPED = MI.1.2[J]
MW[12][J]=0.0
IF(I<>J && IMPED>0.0 ) ; Skip Intrazonal AND Unobserved Cells
MW[12][J] = K2 * ((GEN^AL2) * (ATT^BT2) /(IMPED^(-GM2) ) )
ENDIF
;
GEN = P3[I]
ATT = A3[J]
IMPED = MI.1.2[J]
MW[13][J]=0.0
IF(I<>J && IMPED>0.0 ) ; Skip Intrazonal AND Unobserved Cells
MW[13][J] = K3 * ((GEN^AL3) * (ATT^BT3) /(IMPED^(-GM3) ) )
ENDIF
;
GEN = P4[I]
ATT = A4[J]
IMPED = MI.1.2[J]
MW[14][J]=0.0
IF(I<>J && IMPED>0.0 ) ; Skip Intrazonal AND Unobserved Cells
MW[14][J] = K4 * ((GEN^AL4) * (ATT^BT4) /(IMPED^(-GM4) ) )
ENDIF
;
GEN = P5[I]
ATT = A5[J]
IMPED = MI.1.2[J]
MW[15][J]=0.0
IF(I<>J && IMPED>0.0 ) ; Skip Intrazonal AND Unobserved Cells
MW[15][J] = K5 * ((GEN^AL5) * (ATT^BT5) /(IMPED^(-GM5) ) )
ENDIF
;
GEN = P6[I]
ATT = A6[J]
IMPED = MI.1.2[J]
MW[16][J]=0.0
IF(I<>J && IMPED>0.0 ) ; Skip Intrazonal AND Unobserved Cells

A3-8
 The Study on Urban Transport Master Plan and Feasibility Study in HCM Metropolitan Area (HOUTRANS)
Final Report
VOLUME 4: Transport Database

MW[16][J] = K6 * ((GEN^AL6) * (ATT^BT6) /(IMPED^(-GM6) ) )

ENDIF
;
GEN = P7[I]
ATT = A7[J]
IMPED = MI.1.2[J]
MW[17][J]=0.0
IF(I<>J && IMPED>0.0 ) ; Skip Intrazonal AND Unobserved Cells
MW[17][J] = K4 * ((GEN^AL7) * (ATT^BT7) /(IMPED^(-GM7) ) )
ENDIF
;
ENDJLOOP
;
;
; Set Intra-Zonals
;
MW[11][I] = 0
MW[12][I] = 0
MW[13][I] = 0
MW[14][I] = 0
MW[15][I] = 0
MW[16][I] = 0
MW[17][I] = 0
;
;
;
;
ENDRUN

Step 2: Final Step Distributed Trip Matrices Models

In the second step the initially synthesized trips are controlled to tripend model output trip
generations and attractions using Fratar method. Once this process is completed, Intrazonal trips
are estimated and added to each of the seven trip matrices processed in a single run of the
FRATAR program run.

Program Used: FRATAR

Input Files:

1. Initial Trip Distributed Trip Matrix with Seven Trip tables for a Single Trip Purpose File;

2. Trip Generations and Attractions as P1 to P& & A1 to A7 for each of the Seven HH
categories output by the Tripend Model.

3. Proportion of Intrazonal Trips DBF File

Output File:
1. Out Internal <> Internal Trip Matrix File Containing seven trip tables , one for each
household categories for a single trip purpose. Including Intrazonal Trips

Step-2 Script File for FRATAR Program:

A3-9
 The Study on Urban Transport Master Plan and Feasibility Study in HCM Metropolitan Area (HOUTRANS)
Final Report
VOLUME 4: Transport Database

; Do not change filenames or add or remove FILEI/FILEO statements using an editor.

Use Cube/Application Manager.
RUN PGM=FRATAR PRNFILE="C:\HOUTRANS\MODEL_S3\HDFRA00A.PRN"
MSG='Work_Trip_Dist'
FILEI ZDATI[2] = "C:\HOUTRANS\MODEL_S3\INPUTDATA\PR_INTRAS.DBF"
;
FILEI ZDATI[1] = "C:\HOUTRANS\MODEL_S3\WRK_TEND.DAT"
;
;
FILEI MATI[1] = "C:\HOUTRANS\MODEL_S3\HDMAT00A.MAT"
;
FILEO MATO[1] = "C:\HOUTRANS\MODEL_S3\WRK_DIST.MAT",
DEC=2,
MO=1-7 NAME = WD_CAROW,WD_CARNO, WD_2MCOW,WD_2MCNO, WD_1MCOW,WD_1MCNO,
WD_NOVEH
;
PARAMETERS MAXITERS=30, MAXPURPS=7, MAXRMSE=0.01, ZONES=270
;
;
SETPA P[1]=(P1-MIN(P1,A1)*PR_INT20), A[1]=(A1-MIN(P1,A1)*PR_INT20),
MW[1]=MI.1.1, CONTROL=P
SETPA P[2]=(P2-MIN(P2,A2)*PR_INT20), A[2]=(A2-MIN(P2,A2)*PR_INT20),
MW[2]=MI.1.2, CONTROL=P
SETPA P[3]=(P3-MIN(P3,A3)*PR_INT20), A[3]=(A3-MIN(P3,A3)*PR_INT20),
MW[3]=MI.1.3, CONTROL=P
SETPA P[4]=(P4-MIN(P4,A4)*PR_INT20), A[4]=(A4-MIN(P4,A4)*PR_INT20),
MW[4]=MI.1.4, CONTROL=P
SETPA P[5]=(P5-MIN(P5,A5)*PR_INT20), A[5]=(A5-MIN(P5,A5)*PR_INT20),
MW[5]=MI.1.5, CONTROL=P
SETPA P[6]=(P6-MIN(P6,A6)*PR_INT20), A[6]=(A6-MIN(P6,A6)*PR_INT20),
MW[6]=MI.1.6, CONTROL=P
SETPA P[7]=(P7-MIN(P7,A7)*PR_INT20), A[7]=(A7-MIN(P7,A7)*PR_INT20),
MW[7]=MI.1.7, CONTROL=P
;
MW[1][I] = MIN(P1,A1)*PR_INT20[I]
MW[2][I] = MIN(P2,A2)*PR_INT20[I]
MW[3][I] = MIN(P3,A3)*PR_INT20[I]
MW[4][I] = MIN(P4,A4)*PR_INT20[I]
MW[5][I] = MIN(P5,A5)*PR_INT20[I]
MW[6][I] = MIN(P6,A6)*PR_INT20[I]
MW[7][I] = MIN(P7,A7)*PR_INT20[I]
;
ENDRUN

Distribution Model Output Trip Matrices:

There are 35 (5 trip purposes x 7 HH categories)) trip matrices output by the trip Distribution model.
These matrices are then input to the Modal Split Model. The last Step in the Figure (MATRIX
program with execution order 11, shows how all matrices may be combined for analyses.

A3-10
 The Study on Urban Transport Master Plan and Feasibility Study in HCM Metropolitan Area (HOUTRANS)
Final Report
VOLUME 4: Transport Database

3.3 MODAL SPLIT MODELS

Build Time & Distance Skims Matrices
The key input to the modal split models is the zone-to-zone travel time and distance skims. These
are derived from the network model. The complete methodology used for the derivation of time
and distance skims has been described in Annex 2.1. The only difference is that the Input
network file represents the future year network under test. The remaining process is the same,
and is not repeated here.

Modal Split Model General Description

The modal split model is a 2-Step process, as described in Section 3.4 and illustrated in Figure
3.2, and figures on the next page show the implementation under CUBE/Voyager Programs. In the
fist step the “Walk” only trips are estimated and subtracted from the total distributed trip matrices.
In the second step the remaining non-walk “mechanized” trips are split between competing modes.
Both steps are executed for each trip purpose as the modal split parameters differ by trip purpose.

Figures on the next page illustrate:

1. Step 1 of walk modal split for each of the five trip purposes executed using program
MATRIX as programs with execution order 1-5.
2. Program MATIRX runs with execution order 6 to 9 are for analyses of the output matrices
and are not necessary for the model run.

3. Three sub-groups, one for each of the three modal split scenarios. Each sub-group runs
the same modal split modal, what differs is the cost of travel by each mode, input to the
modal split model giving different modal split trip matrices under each scenario. For details
see Section 3.4.

4. The execution order for each sub-group has been set to zero, so that the any desired
scenario may be executed as required by the test under consideration.
5. Run of any one of the three sub-group modal split scenarios will produce trips by purpose
and mode for further analyses and as input to peak period models prior to assignment.

6. Program MATRIX with execution orders from 10-12 are for analyses of final modal split
matrices by mode and purpose only and are not necessary for the next stage – Peak
period models.

7. However, all matrices are also added to form daily trip matrices by mode and purpose for
analyses of trips as shown in the second figure on the next page for the Base Case modal
split scenario.

A3-11
 The Study on Urban Transport Master Plan and Feasibility Study in HCM Metropolitan Area (HOUTRANS)
Final Report
VOLUME 4: Transport Database

Overall Modal Split Model – Walk and Non-walk Split

Mai Modal Split – Base Case Scenario

A3-12
 The Study on Urban Transport Master Plan and Feasibility Study in HCM Metropolitan Area (HOUTRANS)
Final Report
VOLUME 4: Transport Database

Step-1 Walk Modal Split Model

In this Step walk trips are estimated using Program MATRIX. The estimated walk mode trips are
then subtracted from the total distributed trip matrices. The outputs are trips by walk and all other
non-walk mode trips combined. The process is executed five times, once for each of the five trip
purposes, with parameters calibrated for each purpose and as given Table 3.7. The following
example is for trip purpose Work only.

Program Used: MATRIX

Input Files:

1. Time and distance skims

2. Initial Trip Distributed Trip Matrix with Seven Trip tables (by household category) for a
Single Trip Purpose File;
3. Modal Split Model zonal definition file for the estimation of intra-zonal distances from the
zonal area.

Output File:
1. Out Internal <> Internal trip matrix file of walk trips by all household categories combined.

2. Out Internal <> Internal trip matrix file containing seven trip tables, one for each household
category, for a single trip purpose for all modes of travel, including Intra-zonal Trips.

Step-1 Script File for MATRIX Program for Trip Purpose Work:
; Do not change filenames or add or remove FILEI/FILEO statements using an editor.
Use Cube/Application Manager.
RUN PGM=MATRIX PRNFILE="C:\HOUTRANS\MODEL_S3\HMMAT00A.PRN" MSG='Work MS_Walk'
FILEI MATI[1] = "C:\HOUTRANS\MODEL_S3\TIME&DIST.MAT"
FILEI ZDATI[1] = "C:\HOUTRANS\MODEL_S3\INPUTDATA\MS20_PAR.DBF"
FILEI MATI[2] = "C:\HOUTRANS\MODEL_S3\WRK_DIST.MAT"
;
FILEO MATO[1] = "C:\HOUTRANS\MODEL_S3\WRK_WALK.MAT",
Dec=2,
MO=33 NAME = WRK_WLK_T
;
FILEO MATO[2] = "C:\HOUTRANS\MODEL_S3\HMMAT00A.MAT",
DEC=2,
MO=21-27 NAME = W_Mot_Car_O, W_Mot_Car_N, W_Mot_2MC_O, W_Mot_2MC_N,
W_Mot_1MC_O, W_Mot_1MC_N, W_Mot_Non_O
;
PARAMETERS ZONES=270
;
; COMPUTE Walk TRIPS
;
MW[1] = MI.1.2 ; Set Dsitance Matrix
MW[1][I] = RADIUS[I] ; Set Intra Distances=SQRT of The Zonal Area.
;
;
JLOOP ;
MW[2][J]=0.0
IF(MW[1][J] < 10.0) MW[2][J] = 1 / ( 1 + EXP( 0.3645 * MW[1][J] + 2.0755 )) ;
Prob of Walk Trips for Veh_Owners

A3-13
 The Study on Urban Transport Master Plan and Feasibility Study in HCM Metropolitan Area (HOUTRANS)
Final Report
VOLUME 4: Transport Database

;
MW[3][J]=0.0
IF(MW[1][J] < 10.0) MW[3][J] = 1 / ( 1 + EXP( 0.2892 * MW[1][J] + 0.2007 )) ;
Prob of Walk Trips for Non_Veh_Owners
;
; Set max Walk Distance=5km
;
IF(MW[1][J] >5.0) MW[2][J]=0.0
IF(MW[1][J] >5.0) MW[3][J]=0.0
;
ENDJLOOP
;
;
; Calculate Walk Trips
;
MW[11] = MI.2.1 * MW[2] ; Car Owner Matrix Walk Trips
MW[12] = MI.2.2 * MW[3] ; Car Non_Owner Matrix Walk Trips
;
MW[13] = MI.2.3 * MW[2] ; 2_MCY Owner Matrix Walk Trips
MW[14] = MI.2.4 * MW[3] ; 2_MCY Non_Owner Matrix Walk Trips
;
MW[15] = MI.2.5 * MW[2] ; 1_MCY Owner Matrix Walk Trips
MW[16] = MI.2.6 * MW[3] ; 1_MCY Non_Owner Matrix Walk Trips
;
MW[17] = MI.2.7 * MW[3] ; Non_Owner Matrix Walk Trips
;
; Calculate Motorised= Non_Walk Trips
;
MW[21] = MI.2.1 - MW[11] ; Car Owner Matrix Motorised Trips
MW[22] = MI.2.2 - MW[12] ; Car Non_Owner Matrix Motorised Trips
;
MW[23] = MI.2.3 - MW[13] ; 2_MCY Owner Matrix Motorised Trips
MW[24] = MI.2.4 - MW[14] ; 2_MCY Non_Owner Matrix Motorised Trips
;
MW[25] = MI.2.5 - MW[15] ; 1_MCY Owner Matrix Motorised Trips
MW[26] = MI.2.6 - MW[16] ; 1_MCY Non_Owner Matrix Motorised Trips
;
MW[27] = MI.2.7 - MW[17] ; Non_Owner Matrix Motorised Trips
;
;
; Compute Walk Trips by Ownership
;
MW[31] = MW[11] + MW[13] + MW[15] ; Wlak Trips by Veh Owners
MW[32] = MW[12] + MW[14] + MW[16] + MW[17] ; Wlak Trips by Non_Veh Owners
;
MW[33] = MW[31] + MW[32] ; Total Walk trips
;
;
;
;
ENDRUN

The non-walk output trip matrices are the input to the main modal split described next.

A3-14
 The Study on Urban Transport Master Plan and Feasibility Study in HCM Metropolitan Area (HOUTRANS)
Final Report
VOLUME 4: Transport Database

Step-2 Non-Walk Modal Split Model

In this Step, initially Generalized Cost (GC) of travel for each mode is estimated according to the
scenario under consideration, and trip purpose being modeled, according to the calibrated model
form. The generalized costs of travel required, by each mode and for each trip purpose are
identified in Table 3.8, i.e. parameters with non-zero value. The program used for the calculation
of GC is MATRIX. Input to the program are base case time and distance skims and zonal
information required for the input of certain zonal costs as illustrated by the script file.

1. Calculation of Generalized Costs of Travel for Modal Split Using Program: MATRIX
Input Files:

1. Time and distance skims

2. Modal Split Model zonal attribute file for the estimation of intra-zonal costs and other zone
specific costs i.e. Parking, Area License Fee & Bus Fares for Public Transport.

Output File:

A matrix file containing following 13 Inter and intra zonal Tables generalized costs (GC) and
other attributes of travel by each mode. The tables are:

1. Travel Time at Free Flow Speed

2. Travel Distance

3. Free Flow (FF) Speed; Intra-zonal FF speed assumed to 40kph, and later adjusted
according to mode and zone location.

4. Travel Time by Bicycle @ 10kph

5. Travel Speed by bicycle set to 10kph.

6. Travel Time by Motorcycle @ Speed as given in Table 7, described next.

7. Travel Speed by Motorcycle.

8. Travel Time by Car and Bus @ Speed as given in Table 9, described next.

9. Travel Speed by Car or Bus (assumed to be the same).

10. Bicycle (VoC) operation cost.

11. Motorcycle (VoC) operating cost.

12. Car (VoC) Operating Cost.

13. Bus Fares.

The above process was used to calculate generalized cost for Modal Split for Base Case,
Scenario 1 and Scenario 2. The Vehicle Operating Cots (VoC) used for each scenario are given
below for travel speeds between 5 & 90 kph. In base case petrol prices are assumed to remain
constant at today’s prices in real terms. Under Scenario 1 fuel prices are assumed to go up by
50% in real terms between now and 2020. For Scenario 2 the fuel prices are assumed to be 100%
higher than today by 2020. These price rises are assumed to reflect future policies to deter
motorists from using the private mode, and transfer to PT. Other out-of-pocket charges are also

A3-15
 The Study on Urban Transport Master Plan and Feasibility Study in HCM Metropolitan Area (HOUTRANS)
Final Report
VOLUME 4: Transport Database

assumed to be higher for private mode between base case and the two scenarios. These are
detailed in tables below and used in the cost calculation program.

Table for GC by Mode for Each Scenario

GC Base Case VND/km Scenario 1 VND/km Scenario 2 VND/km
Description
Bicycle M/C Car Bicycle M/C Car Bicycle M/C Car

VoC @ 5kph 1,761 7,301 1,844 7,910 1,927 8,519

VoC @ 10kph 1,020 4,328 1,074 4,726 1,128 5,123
VoC @ 20kph 629 2,756 668 3,043 707 3,329
VoC @ 30kph 494 2,209 528 2,458 563 2,708
VoC @ 40kph 415 1,891 447 2,121 479 2,351
57.0 57.0 57.0
VoC @ 50kph 377 1,747 407 1,971 439 2,196
VoC @ 60kph 370 1,741 403 1,973 436 2,206
VoC @ 70kph 377 1,781 409 2,027 443 2,274
VoC @ 80kph 386 1,848 422 2,114 459 2,379
VoC @ 90kph 403 1,955 443 2,248 484 ,540
Parking / Trip 250 1,000 2,000 500 1,500 2,000 500 2,000 2,000
Area
0 0 0 0 0 0 0 3,500 7,000
License Fee
Notes:
[1] Bicycle speed was set to 10 kph for all areas for all purposes.

[2] Motorcycle and Car (Bus Speeds assumed= Car Speed) speeds were estimated according to
the travel between different areas. All Speeds were set to minimum of 5kph, and maximum of
90kph. VoC for speed between those specified above were estimated using linear interpolation.
The area to area speed factors by mode of travel applied to the free-flow speed are given below.

Area to Area Speed Factors Applied to FF Speed to Get Operating Speeds for Modal Split
Mode Motorcycle Speed Factors Car & Bus Speed Factors

Area Area-1 Area-2 Area-3 Area-1 Area-2 Area-3

Area-1 0.7 0.7 0.7 0.6 1.0 1.0

Area-2 0.7 0.7 0.8 1.0 1.0 1.0
Area-3 0.7 0.8 0.8 1.0 1.0 1.0

Area-1: Zones 1-136 (All HCMC Inner Area Zones)

Area-2: Zones 137-209 (All HCMC Outer Area Zones)

Area-3: Zones 210-270 (All Rural Area Zones)

Bus Fare was estimated as VND1,000 for first 4km, and VND250/km thereafter for all scenarios.

A3-16
 The Study on Urban Transport Master Plan and Feasibility Study in HCM Metropolitan Area (HOUTRANS)
Final Report
VOLUME 4: Transport Database

Step-1 Script File for MATRIX Program for ALL Trip Purposes Base Case Modal Split:
; Do not change filenames or add or remove FILEI/FILEO statements using an editor.
Use Cube/Application Manager.
RUN PGM=MATRIX PRNFILE="C:\HOUTRANS\MODEL_S3\HBMAT00A.PRN" MSG='Calc_GC_Base
Case Fuel*1'
FILEI MATI[1] = "C:\HOUTRANS\MODEL_S3\TIME&DIST.MAT"
;
FILEI ZDATI[1] = "C:\HOUTRANS\MODEL_S3\INPUTDATA\MS20_PAR.DBF"
;
;
FILEO MATO[1] = "C:\HOUTRANS\MODEL_S3\HBMAT00A.MAT",
DEC=2,
MO=1-9,10-13 NAME = TIME, DIST, FFSPD, BCY_T, BCY_S, MCY_T, MCY_S, C&B_T, C&B_S,
BCY_VoC, MCY_VoC, Car_VoC, Bus_Fare

;
;
PARAMETERS ZONES=270
;
;
;
BCY_PARK = 250
MCY_PARK = 1000
CAR_PARK = 2000
;
;
Parkfac=0 ; Normal Parking, Set=1 If Extra Parking in CBD (Dist 1,3,5&10)
Destination Zones to be Applied
;
ALFac=0 ; Area Licence Fee, Set=1 If ALF to be Applied for Trips from Outside
to HCMC (Zones 1-136).
MCY_ALF = 3500
CAR_ALF = 7000
;
FFFAC=0 ; SET = 1 IF Flat Bus Fare in HCMC Area (Zones 1-136) to be Applied
FLAT_FARE = 1000
;
; Set Up Lookup Table for MCY & Car VoCs for OP_Speeds=5~90kph to be Inyterpolated
according to Speed
;
LOOKUP NAME=VOPCOST, ; Lookup Table
LOOKUP[1]=1, RESULT=2, ; Curve 1 for MCY Speed in Field_1, VoC in
Field_2.
LOOKUP[2]=1, RESULT=3, ; Curve 2 for Car Speed in Field_1, VoC in
Field_3.
;
FAIL=0,0,9999, INTERPOLATE=Y,
;
; SPD V_Mc V_Car VoC Table for Base Case
R=' 5 1761 7301', ; VoC of MCY & Car for Speed=5kph
' 10 1020 4328',
' 20 629 2756',
' 30 494 2209',
' 40 415 1891',
' 50 377 1747',
' 60 370 1741',
' 70 377 1781',
' 80 386 1848',
' 90 403 1955' ; End of VoC data by Speed
;

A3-17
 The Study on Urban Transport Master Plan and Feasibility Study in HCM Metropolitan Area (HOUTRANS)
Final Report
VOLUME 4: Transport Database

;
; Calculate Free Flow Speed
;
MW[1]=MI.1.1 ; Set FREE Flow Time
MW[2]=MI.1.2 ; Set Distance Matrix
;
; COMPUTE Intra Zonal Distance
;
INTRA_DIST = RADIUS[I] ; Set Intra Distances.
MW[2][I]= INTRA_DIST
;
IF(I=266) MW[2][I] = 2.0 ; Set External Intrazonal Dsitances = 2.0km
IF(I=267) MW[2][I] = 2.0 ; to Avoid Divide by Zero in Calcs
IF(I=268) MW[2][I] = 2.0
IF(I=269) MW[2][I] = 2.0
IF(I=270) MW[2][I] = 2.0
;
;
; Compute Free Flow Speed - for Intra-Zonal = 40kph
;
MW[3]=0.0 ; Initialis
MW[3][I]=40.0 ; Set Intra-Zonal FREE FLOW Speed = 40kph
;
IF(MW[3][I]>0) MW[1][I] = (MW[2][I] / MW[3][I]) * 60.0 ; Calculate Intra-Zonal
Time
;
JLOOP
IF(MW[1][J]>0) MW[3][J]=MW[2][J]/(MW[1][J]/60.0)
ENDJLOOP
;
; Calculate BCY Speed and Time
;
MW[5] = 10.0 ; Fixed as 10kph
MW[4] = (MW[2] / MW[5]) * 60.0 ; BCY Time in Mins
;
; Calculate BCY VoC @ 10kph & Parking
Charges Occ=1.10
;
JLOOP
MW[10][J] = ( 57.0 * MW[2][J] + BCY_PARK + BCY_PARK * Parkfac * PARK1[J] )
/ 1.10
ENDJLOOP
;
;
; Calculate Area to Area Speed Factor for MCY and Car (Bus=Car)
;
LOOKUP NAME=OPSP_FAC, ; Lookup Table
LOOKUP[1]=1, RESULT=2, ; Curve 1 for Area in Field_1, MCY Speed Factor
in Field_2.
LOOKUP[2]=1, RESULT=3, ; Curve 2 for Area in Field_1, CAR=BUS Speed
Factor in Field_2.
;
FAIL=0,0,9999, INTERPOLATE=N,
;
; AREA MCY_SPF CAR_SPF
R=' 1 0.7 0.6 ', ; Factor of MCY & Car=Bus
' 2 0.7 1.0 ',
' 3 0.7 1.0 ',
' 4 0.7 1.0 ',
' 5 0.8 1.0 ',
' 6 0.8 1.0 '

A3-18
 The Study on Urban Transport Master Plan and Feasibility Study in HCM Metropolitan Area (HOUTRANS)
Final Report
VOLUME 4: Transport Database

; End of SPEED FACTOR DATA

;
JLOOP
;
IIZ = SPD_AREA[I]
JJZ = SPD_AREA[J]
;
IF(IIZ=1 && JJZ=1) KKZ=1
IF(IIZ=1 && JJZ=2) KKZ=2
IF(IIZ=1 && JJZ=3) KKZ=3
;
IF(IIZ=2 && JJZ=1) KKZ=2
IF(IIZ=2 && JJZ=2) KKZ=4
IF(IIZ=2 && JJZ=3) KKZ=5
;
IF(IIZ=3 && JJZ=1) KKZ=3
IF(IIZ=3 && JJZ=2) KKZ=5
IF(IIZ=3 && JJZ=3) KKZ=6
;
MCY_SPD_FACT = OPSP_FAC(1,KKZ)
CAR_SPD_FACT = OPSP_FAC(2,KKZ)
;
MW[7][J] = MW[3][J] * MCY_SPD_FACT ; Assumed as FF Speed * MCY_SPD_FACTOR
MW[9][J] = MW[3][J] * CAR_SPD_FACT ; Assumed as CAR_SPD_FACTOR * FF Speed
;
ENDJLOOP ; END of Selection of Speed Factor
;
; Calculate MCY Speed and Time & VoC
;
JLOOP
;
IF( MW[7][J]< 5.0 ) MW[7][J] = 5.0 ; Set Min Speed= 5.0
IF( MW[7][J]>90.0 ) MW[7][J] =90.0 ; Set Max Speed=90.0
IF(MW[7][J]>0.0) MW[6][J] = (MW[2][J] / MW[7][J]) * 60.0
;
ENDJLOOP
;
JLOOP ; Compute MCY VoC
;
OPSPD = MW[7][J] ; Get Operating Speed
MCYVOC = VOPCOST(1,OPSPD) ; Get VoC at Operating Speed
MW[11][J] = MCYVOC * MW[2][J] ; Set MW[11] = VoC
;
MW[11][J] = ( MW[11][J] + MCY_PARK + MCY_PARK * Parkfac * PARK1[J] ) ; Add
Parking Charges at Destination Zone
IF (ALF1[I]=0 && ALF1[J]=1) MW[11][J] = MW[11][J] + (MCY_ALF * ALFac ) ;
Add ALF for Trips to HCMC.
;
MW[11][J] = MW[11][J] / 1.50 ; Calculate Cost Per Person Occ=1.50
;
ENDJLOOP
;
; Calculate Car Speed and Time & VoC
;
JLOOP
IF( MW[9][J]< 5.0 ) MW[9][J] = 5.0 ; Set Min Speed= 5.0
IF( MW[9][J]>90.0 ) MW[9][J] =90.0 ; Set Max Speed=90.0
IF(MW[9][J]>0.0)MW[8][J] = (MW[2][J] / MW[9][J]) * 60.0
ENDJLOOP
;
JLOOP ; Compute Car VoC

A3-19
 The Study on Urban Transport Master Plan and Feasibility Study in HCM Metropolitan Area (HOUTRANS)
Final Report
VOLUME 4: Transport Database

;
OPSPD = MW[9][J] ; Get Operating Speed
CARVOC = VOPCOST(2,OPSPD) ; Get VoC at Operating Speed
MW[12][J] = CARVOC * MW[2][J] ; Set MW[12] = VoC
;
MW[12][J] = ( MW[12][J] + CAR_PARK + CAR_PARK * Parkfac * PARK1[J] ) ; Add
Parking Charges at detination Zone
IF (ALF1[I]=0 && ALF1[J]=1) MW[12][J] = ( MW[12][J] + CAR_ALF * ALFac ) ;
Add ALF for Trips to HCMC.
;
MW[12][J] = MW[12][J] / 2.00 ; Calculate Cost Per Person Occ=2.00
;
ENDJLOOP
;
; Calculate Bus Fare Based on 1000VN$ (for Ist 4km) + 250/km
Distance
;
;
JLOOP ; Compute Bus Fare
;
MW[13][J] = 1000.0 ; Set Initial Bus Fare for Up to 4km

;
IF (MW[2][J] > 4.0) MW[13][J] = MW[13][J] + (250 * ( MW[2]-4.0 )) ; Add FARE
For Dist over 4km.
IF (FFFAC=1 && FFARE1[I]=1 && FFARE1[J]=1) MW[13][J] = FLAT_FARE ; Set HCMC
Area Fare=Flat Fare of VN$1,000.
;
ENDJLOOP
;
;
ENDRUN

2. Modal Split by Trip Purpose

Main modal split model is run five times, once for each trip purpose as shown above in Base Case
Modal Split Figure (execution order 2-6). The MATRIX program is used to calculate probability of
mode choice, and the probabilities are applied to distributed trip matrices by trip purpose and
household categories. The script file below is for trip purpose “Work” and all seven household
categories, and for execution order 2. The other Script files are not given here. Similarly the use of
program Matrix after the modal split at execution order 7-9 is simply to aggregate matrices by
mode of travel for assignment and other analyses.

The Modal Split model parameters for each trip purpose and by HH category are given in Section
3.4.

Using Program: MATRIX

Input Files:
1. Generalized Cost skims Produced at above step.
2. Distributed Trip Matrix file for one purpose (Work in this case) with seven trip tables, one
for each household categories.

A3-20
 The Study on Urban Transport Master Plan and Feasibility Study in HCM Metropolitan Area (HOUTRANS)
Final Report
VOLUME 4: Transport Database

Output File:

A matrix file containing 4 Inter and intra zonal Trip Tables for travel by Bicycle, Motorcycle,
Car and PT (Bus & MTR) for all seven household categories combined, and for trip purpose
work.

Step-2 Script File for Program MATRIX for Work Trip Purposes Base Case Modal Split:
; Do not change filenames or add or remove FILEI/FILEO statements using an editor.
Use Cube/Application Manager.
RUN PGM=MATRIX PRNFILE="C:\HOUTRANS\MODEL_S3\HBMAT00B.PRN" MSG='Work Main
M-Split'
FILEI MATI[2] = "C:\HOUTRANS\MODEL_S3\HMMAT00A.MAT"
FILEI MATI[1] = "C:\HOUTRANS\MODEL_S3\HBMAT00A.MAT"
;
;
;
FILEO MATO[1] = "C:\HOUTRANS\MODEL_S3\HBMAT00B.MAT",
DEC=4*2,
MO=71-74 NAME=WRK_BCY, WRK_MCY, WRK_CAR, WRK_BUS ; WORK
;
; Set VoC
;
MW[9] = MI.1.2 ; Distance
MW[1] = MI.1.10 ; BCY_VoC
MW[2] = MI.1.11 ; MCY_VoC
MW[3] = MI.1.12 ; Car_VoC
MW[4] = MI.1.13 ; Bus_Fare
MW[5] = MI.1.4 ; BCY_Time
;
MW[61] = MI.2.1 ; Car-HH & Avail=1 Trips
MW[62] = MI.2.2 ; Car-HH & Avail=0 Trips
MW[63] = MI.2.3 ; 2MC-HH & Avail=1 Trips
MW[64] = MI.2.4 ; 2MC-HH & Avail=0 Trips
MW[65] = MI.2.5 ; 1MC-HH & Avail=1 Trips
MW[66] = MI.2.6 ; 1MC-HH & Avail=0 Trips
MW[67] = MI.2.7 ; NMV-HH & Avail=0 Trips
;
;
; Model Spli Model - Purpose = Work Person Trips Only
;
;
; SET - PARAMETERS - BCY
;
DFACT = -0.15
;
BCYK = 0.70
;
BVOC = -0.0069
BCHH = 0.0
B2MC = 0.0
B1MC = +1.7450
BNMH = +2.3793
BOWN = -0.9700
;
; SET - PARAMETERS - MCY
;
MCYK = -0.26
;

A3-21
 The Study on Urban Transport Master Plan and Feasibility Study in HCM Metropolitan Area (HOUTRANS)
Final Report
VOLUME 4: Transport Database

MVOC = -0.0010
MCHH = 0.0
M2MC = +1.1726
M1MC = +1.5450
MNMH = 0.0
MOWN = +2.9246
;
; SET - PARAMETERS - CAR
;
CARK = 0.0
;
CVOC = -0.00030
CCHH = +2.2974
C2MC = 0.0
C1MC = 0.0
CNMH = 0.0
COWN = 0.0
;
; SET - PARAMETERS - BUS
;
BUSK = 0.3250
;
PVOC = -0.00096
PCHH = 0.0
P2MC = -1.2991
P1MC = -0.9346
PNMH = 0.0
POWN = 0.0
;
;
; Compute Utility Mxs for - BCY
;
; = Const Dist-rel VoC_by BCY Car-HH 2MCY-HH
1-MCY-HH NM-HH Available=
MW[11] = BCYK + DFACT*MW[9] + BVOC * MW[1] + BCHH*1
+BOWN * 1 ; Car-HH & Avail=1
MW[12] = BCYK + DFACT*MW[9] + BVOC * MW[1] +
BCHH*1 ; Car-HH & Avail=0
MW[13] = BCYK + DFACT*MW[9] + BVOC * MW[1] + B2MC * 1
+BOWN * 1 ; 2MC-HH & Avail=1
MW[14] = BCYK + DFACT*MW[9] + BVOC * MW[1] + B2MC *
1 ; 2MC-HH & Avail=0
MW[15] = BCYK + DFACT*MW[9] + BVOC * MW[1] + B1MC * 1
+BOWN * 1 ; 1MC-HH & Avail=1
MW[16] = BCYK + DFACT*MW[9] + BVOC * MW[1] + B1MC *
1 ; 1MC-HH & Avail=0
MW[17] = BCYK + DFACT*MW[9] + BVOC * MW[1]
+ BNMH*1 ; NMV-HH & Avail=0
;
;
; Compute Utility Mxs for - MCY
;
; = Const VoC_by MCY Car-HH 2MCY-HH 1-MCY-HH NM-HH
Available=
MW[21] = MCYK + MVOC * MW[2] + MCHH * 1 +
MOWN * 1 ; Car-HH & Avail=1
MW[22] = MCYK + MVOC * MW[2] + MCHH *
1 ; Car-HH & Avail=0
MW[23] = MCYK + MVOC * MW[2] + M2MC * 1 +
MOWN * 1 ; 2MC-HH & Avail=1

A3-22
 The Study on Urban Transport Master Plan and Feasibility Study in HCM Metropolitan Area (HOUTRANS)
Final Report
VOLUME 4: Transport Database

MW[24] = MCYK + MVOC * MW[2] + M2MC *

1 ; 2MC-HH & Avail=0
MW[25] = MCYK + MVOC * MW[2] + M1MC * 1 +
MOWN * 1 ; 1MC-HH & Avail=1
MW[26] = MCYK + MVOC * MW[2] + M1MC *
1 ; 1MC-HH & Avail=0
MW[27] = MCYK + MVOC * MW[2] + MNMH *
1 ; NMV-HH & Avail=0
;
;
; Compute Utility Mxs for - CAR
;
; = Const VoC_by CAR Car-HH 2MCY-HH 1-MCY-HH NM-HH
Available=
MW[31] = CARK + CVOC * MW[3] + CCHH * 1 +
COWN * 1; Car-HH & Avail=1
MW[32] = CARK + CVOC * MW[3] + CCHH *
1 ; Car-HH & Avail=0
MW[33] = CARK + CVOC * MW[3] + C2MC * 1 +
COWN * 1; 2MC-HH & Avail=1
MW[34] = CARK + CVOC * MW[3] + C2MC *
1 ; 2MC-HH & Avail=0
MW[35] = CARK + CVOC * MW[3] + C1MC * 1 +
COWN * 1; 1MC-HH & Avail=1
MW[36] = CARK + CVOC * MW[3] + C1MC *
1 ; 1MC-HH & Avail=0
MW[37] = CARK + CVOC * MW[3] + CNMH *
1 ; NMV-HH & Avail=0
;
;
; Compute Utility Mxs for - BUS
;
; = Const BUS-FARE Car-HH 2MCY-HH 1-MCY-HH NM-HH
Available=
MW[41] = BUSK + PVOC * MW[4] + PCHH * 1
+POWN * 1 ; Car-HH & Avail=1
MW[42] = BUSK + PVOC * MW[4] + PCHH *
1 ; Car-HH & Avail=0
MW[43] = BUSK + PVOC * MW[4] + P2MC * 1
+POWN * 1 ; 2MC-HH & Avail=1
MW[44] = BUSK + PVOC * MW[4] + P2MC *
1 ; 2MC-HH & Avail=0
MW[45] = BUSK + PVOC * MW[4] + P1MC * 1
+POWN * 1 ; 1MC-HH & Avail=1
MW[46] = BUSK + PVOC * MW[4] + P1MC *
1 ; 1MC-HH & Avail=0
MW[47] = BUSK + PVOC * MW[4] + PNMH *
1 ; NMV-HH & Avail=0
;
;
JLOOP ; SET BCY TRIPS LONGER THAN 120 MINS =0
;
MW[11][J]=EXP(MW[11][J])
MW[12][J]=EXP(MW[12][J])
MW[13][J]=EXP(MW[13][J])
MW[14][J]=EXP(MW[14][J])
MW[15][J]=EXP(MW[15][J])
MW[16][J]=EXP(MW[16][J])
MW[17][J]=EXP(MW[17][J])
;
IF(MW[5][J] > 120.0) MW[11][J]=0.0

A3-23
 The Study on Urban Transport Master Plan and Feasibility Study in HCM Metropolitan Area (HOUTRANS)
Final Report
VOLUME 4: Transport Database

IF(MW[5][J] > 120.0) MW[12][J]=0.0

IF(MW[5][J] > 120.0) MW[13][J]=0.0
IF(MW[5][J] > 120.0) MW[14][J]=0.0
IF(MW[5][J] > 120.0) MW[15][J]=0.0
IF(MW[5][J] > 120.0) MW[16][J]=0.0
IF(MW[5][J] > 120.0) MW[17][J]=0.0
;
ENDJLOOP
;
MW[21] =EXP(MW[21])
MW[22] =EXP(MW[22])
MW[23] =EXP(MW[23])
MW[24] =EXP(MW[24])
MW[25] =EXP(MW[25])
MW[26] =EXP(MW[26])
MW[27] =EXP(MW[27])
;
MW[31] =EXP(MW[31])
MW[32] =EXP(MW[32])
MW[33] =EXP(MW[33])
MW[34] =EXP(MW[34])
MW[35] =EXP(MW[35])
MW[36] =EXP(MW[36])
MW[37] =EXP(MW[37])
;
MW[41] =EXP(MW[41])
MW[42] =EXP(MW[42])
MW[43] =EXP(MW[43])
MW[44] =EXP(MW[44])
MW[45] =EXP(MW[45])
MW[46] =EXP(MW[46])
MW[47] =EXP(MW[47])
;
;
; Calc BCY Probs & Trips & Store Trips
;
MW[51] = MW[11] / ( MW[11] + MW[21] + MW[31] + MW[41] )
MW[52] = MW[12] / ( MW[12] + MW[22] + MW[32] + MW[42] )
MW[53] = MW[13] / ( MW[13] + MW[23] + MW[33] + MW[43] )
MW[54] = MW[14] / ( MW[14] + MW[24] + MW[34] + MW[44] )
MW[55] = MW[15] / ( MW[15] + MW[25] + MW[35] + MW[45] )
MW[56] = MW[16] / ( MW[16] + MW[26] + MW[36] + MW[46] )
MW[57] = MW[17] / ( MW[17] + MW[27] + MW[37] + MW[47] )
;
MW[71] = MW[51]*MW[61] + MW[52]*MW[62] + MW[53]*MW[63] +
MW[54]*MW[64] + MW[55]*MW[65] + MW[56]*MW[66] + MW[57]*MW[67]
;
; Calc MCY Probs & Trips & Store Trips
;
MW[51] = MW[21] / ( MW[11] + MW[21] + MW[31] + MW[41] )
MW[52] = MW[22] / ( MW[12] + MW[22] + MW[32] + MW[42] )
MW[53] = MW[23] / ( MW[13] + MW[23] + MW[33] + MW[43] )
MW[54] = MW[24] / ( MW[14] + MW[24] + MW[34] + MW[44] )
MW[55] = MW[25] / ( MW[15] + MW[25] + MW[35] + MW[45] )
MW[56] = MW[26] / ( MW[16] + MW[26] + MW[36] + MW[46] )
MW[57] = MW[27] / ( MW[17] + MW[27] + MW[37] + MW[47] )
;
;
MW[72] = MW[51]*MW[61] + MW[52]*MW[62] + MW[53]*MW[63] +
MW[54]*MW[64] + MW[55]*MW[65] + MW[56]*MW[66] + MW[57]*MW[67]
;

A3-24
 The Study on Urban Transport Master Plan and Feasibility Study in HCM Metropolitan Area (HOUTRANS)
Final Report
VOLUME 4: Transport Database

; Calc CAR Probs & Trips & Store Trips

;
;
MW[51] = MW[31] / ( MW[11] + MW[21] + MW[31] + MW[41] )
MW[52] = MW[32] / ( MW[12] + MW[22] + MW[32] + MW[42] )
MW[53] = MW[33] / ( MW[13] + MW[23] + MW[33] + MW[43] )
MW[54] = MW[34] / ( MW[14] + MW[24] + MW[34] + MW[44] )
MW[55] = MW[35] / ( MW[15] + MW[25] + MW[35] + MW[45] )
MW[56] = MW[36] / ( MW[16] + MW[26] + MW[36] + MW[46] )
MW[57] = MW[37] / ( MW[17] + MW[27] + MW[37] + MW[47] )
;
;
MW[73] = MW[51]*MW[61] + MW[52]*MW[62] + MW[53]*MW[63] +
MW[54]*MW[64] + MW[55]*MW[65] + MW[56]*MW[66] + MW[57]*MW[67]
;
; Calc BUS Probs & Trips & Store Trips
;
;
MW[51] = MW[41] / ( MW[11] + MW[21] + MW[31] + MW[41] )
MW[52] = MW[42] / ( MW[12] + MW[22] + MW[32] + MW[42] )
MW[53] = MW[43] / ( MW[13] + MW[23] + MW[33] + MW[43] )
MW[54] = MW[44] / ( MW[14] + MW[24] + MW[34] + MW[44] )
MW[55] = MW[45] / ( MW[15] + MW[25] + MW[35] + MW[45] )
MW[56] = MW[46] / ( MW[16] + MW[26] + MW[36] + MW[46] )
MW[57] = MW[47] / ( MW[17] + MW[27] + MW[37] + MW[47] )
;
;
MW[74] = MW[51]*MW[61] + MW[52]*MW[62] + MW[53]*MW[63] +
MW[54]*MW[64] + MW[55]*MW[65] + MW[56]*MW[66] + MW[57]*MW[67]
;
;
;
ENDRUN

A3-25
 The Study on Urban Transport Master Plan and Feasibility Study in HCM Metropolitan Area (HOUTRANS)
Final Report
VOLUME 4: Transport Database

ANNEX 4 HOUTRANS TRAINING COURSE

4.1 HOUTRANS TRAINING COURSE LIST OF PARTICIPENTS

HOUTRANS: CUBE/Voyager Training Courses for TDSI

Staff
Course 2: Friday 13th to Friday 20th February 2004 (6-Days) –
Attendance:
1. Mr Voung Tan Duc - Transport Economist TDSI-South HCMC

2. Mr Tran Minh PHuong - Deputy Chief TDSI-North Hanoi

3. Ms Nguyen Thi Diem Hang - Transport Planner TDSI-North Hanoi

4. Mr Du Khanh Thang - Transport Planner TDSI-North Hanoi

5. Mr Tran Duc Sa - Transport Planner TDSI-North Hanoi

6. Mr Dao Trung Nghia - Transport Planner TDSI-South HCMC

Course 1: Wednesday 14th to Friday 16th January 2004 (3 Days) –

Attendance:
1. Mr Voung Tan Duc - Transport Economist TDSI-South HCMC

2. Mr Nguyen Duy Hung - Transport Planner TDSI-South HCMC

3. Mr Nguyen Khanh Trung - Transport Planner TDSI-South HCMC

A3-26
 The Study on Urban Transport Master Plan and Feasibility Study in HCM Metropolitan Area (HOUTRANS)
Final Report
VOLUME 4: Transport Database

4.2 HOUTRANS TRAINING COURSE HANDOUT

CUBE Training Course

for

TDSI – South & North Staff

13-20 February 2003

By ALMEC Staff

Almec - HOUTRANS 1

Agenda - Overview
† 6 Day Course – Every Day 4-Sessions

† We Will (try) Cover:

¾ Travel Demand Forecasting,

Why Forecast? (Today)
¾ HOUTRANS - Models (Next Week)

Almec - HOUTRANS 2

A3-27
 The Study on Urban Transport Master Plan and Feasibility Study in HCM Metropolitan Area (HOUTRANS)
Final Report
VOLUME 4: Transport Database

Choice and Use of Software in Modelling

¾ HOUTRANS Models
ƒ Level – 1 ( Broad Landuse Model )
ƒ Level – 2 ( “Coarse” Strategic Model )
ƒ Level – 3 ( Disaggregate Strategic Model )

¾ Available Software & Use

ƒ Strada
ƒ Trips
ƒ Others (TP+ / Tranplan / Visim / Emme/2 / Saturn / Paramics)
ƒ Cube Voyager

Almec - HOUTRANS 3

Conceptual Transportation Planning Process

Aims & Objectives

Inventories of existing
conditions - Surveys

Forecasts Land Use &

Travel Demand

Supply – Network

Analysis of Alternatives
-e.g. Roads, LRT, Bus Services
Evaluation of Alternatives
Financial, Economic, Social &
Environmental

Selection & Implementation

Almec - HOUTRANS 4

A3-28
 The Study on Urban Transport Master Plan and Feasibility Study in HCM Metropolitan Area (HOUTRANS)
Final Report
VOLUME 4: Transport Database

Steps in Transport Planning Process

† Surveys & Analysis
„ Land Use & Socio-economic data (HIS, SP, Stats etc)
„ Network Inventory (Roads / Parking etc)
„ Travel Pattern Data – Most Important (HIS, RIS O/D etc)
† Demand Forecasting and Plan Formulation
„ Forecast Travel Demand
„ Supply-side Proposals
† Evaluation of Plans & Policies
„ Test Alternatives
„ Best Schemes / Options which Meet Policy Objectives,
taking into account all relevant considerations ?

Almec - HOUTRANS 5

The Travel Demand Forecasting Process

¾Traffic & Transport Surveys
¾Demand Modelling
¾4-Stage Demand Models
¾Trip Generation/Attraction Models
¾Trip Distribution Models
¾Mode Choice Models
¾Traffic Assignment

Almec - HOUTRANS 6

A3-29
 The Study on Urban Transport Master Plan and Feasibility Study in HCM Metropolitan Area (HOUTRANS)
Final Report
VOLUME 4: Transport Database

4-Stage Modelling Process

Almec - HOUTRANS 7

Notes on 4-Stage Model Process

† Continuous Improvements
† Requires a large amount of data to Build, Calibrate
& Validate Model(s) -
† Common Usage of 4-Stage Models to Prepare
Master Plans for Long Term Strategies &
Proposals, and the Sub-models for Local Schemes
„ Toll road model may use a fixed highway matrix
„ Elasticity may be used to represent impact of traffic
flow from changes in toll levels, for example

Almec - HOUTRANS 8

A3-30
 The Study on Urban Transport Master Plan and Feasibility Study in HCM Metropolitan Area (HOUTRANS)
Final Report
VOLUME 4: Transport Database

Transport Forecasting
– Latest Modelling Techniques
† Current Trends Reflect:
„ Concerns of Decision Makers
„ Availability of Powerful Computing & Tools
„ Desire for Improvements to Modelling Accuracy

† Policy options to be tested can include:

„ New LRT Lines, Fuel Prices, Parking Charges, Area License Fee
„ Road Pricing - Tolls
„ New Bus Routes
„ Other Types of Traffic Control Systems & Demand Management

† Diverse Policy Options are Tested Using a “Coarse” Strategic Model –

HOUTRANS Level-2

Almec - HOUTRANS 9

Transport Forecasting –
Caution on Models
† Models should be seen as simplifications of reality that can
Forecast Estimates of the future situation
† However, they are heavily dependent on the input data and
assumptions
† Should input data and assumptions change between the
time of the original forecasts and the opening of a scheme
(e.g. LRT) then actual traffic would be expected to be
different from the forecast traffic
† Whilst desire may be to build a complex model, it must
always suit the data which is available
Almec - HOUTRANS 10

A3-31