See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/283053170

A Review of Power System State Estimation by Weighted Least Square Technique

Conference Paper · February 2015

CITATIONS READS
3 4,238

2 authors, including:

Ishaq Sheikh
Government Engineering College, Dahod
6 PUBLICATIONS 4 CITATIONS

SEE PROFILE

All content following this page was uploaded by Ishaq Sheikh on 21 October 2015.

The user has requested enhancement of the downloaded file.

 A Review of Power System State
Estimation by Weighted Least Square Technique
C.Vaishnavi Prof. I.A. Sheikh
PG Student,Power System Assistant professor
c.vaishnavi17@gmail.com ishaqsheikh@gcet.ac.in
GCET, V.V.Nagar GCET, V.V.Nagar

Abstract -- The control centers of utilities receive

abundant raw measurements from power system
networks. The measurements may be destroyed or
lost while being transmitted to the control centers
due to defects in sensors or loss of communication
during a natural disaster. In such cases the raw
data has to be processed before being used for
analysis by various Energy Management System
(EMS) applications. State estimation is a computer
program that helps to process the raw
measurements and determine the power system
states from them. Vector consisting of bus voltage
magnitudes and their respective bus phase angles
are states of an electric power system. In recent Figure 1: Functions of State Estimator
years, a wide range of state estimation methods has
emerged, and this paper provides an overview of any corrective or preventive action [1]. It is possible to
the commonly used state estimation method. It get a reduced set of data or corrupted data during
outlines Weighted Least Square Technique and the extreme contingencies for further analysis [1]. State
steps involved in the estimation. Estimation helps to get a better picture of the power
system with an available set of measurements.

I. INTRODUCTION The general concepts of state estimation in power

Nowadays mechanism of data acquisition and system including formulation and implementation
transmission are developed. However small random have been discussed by the authors in [6,7]. The effect
errors are always present in meter readings as real time of massive data loss from a control center in a large
readings will not be perfect. These uncertainties are scale power system and the type of measurements that
due to error in metering, communication errors, can be included to provide a better state estimator
incomplete metering and error in mathematical model solution was analyzed [10].
or unexpected system changes. Measurements are
affected by errors due to bad functioning of measuring Identification of efficient state estimation algorithms
instrument in transmission lines, bad calibration of at different levels of data redundancy is necessary
equipments. Bad data points will occur. Sometimes during extreme conditions. It is also required to
meter readings may be lost completely or destroyed explore and evaluate different state estimation
while being transmitted to control centers due to methods, when there are fewer amounts of data or no
defects in sensors or loss of communication during data from a system network.
natural disasters. Basically, we can define the state estimator as a
So online control of electrical power systems demand calculation program carried in real time with the
the measurement processing to determine true state of purpose of providing a complete, coherent and
system. The function of state estimator is shown in reliable database, which can describe the
figure 1. The most electrical state of the network [11]-[12].
used are those based on the weighted least squares [2].
The exact state of the electric grid is needed to perform
 II. OVERVIEW OF STATE ESTIMATION

Power system network is monitored by many sensors

like current transformer, potential transformer, relay
and Phasor Measurement Units. A conventional state
estimator estimates the optimal voltage magnitudes
and phase angles at the buses of the entire system [13].
The state estimation for a power network involves
gathering the real time measurement data like line
flows, power injections and voltage measurement
through SCADA and calculating state vector using
predefined specific state estimation algorithm. State
estimation helps to estimate the states of the system as
wells as the values of the measurements in presence of
error in acquisition of measured data. The state
estimation results forms the basis for other EMS
functions, like security analysis, optimal load dispatch, Figure 2: Flowchart of state estimation algorithm [2]
voltage stability analysis, etc.
The WLS state estimation program is implemented for
The application of state estimation in power system a 3 Bus system and convergence is checked in [3].First
with sensors located in a substation. The data from power flow of the 3-Bus system is run to obtain
sensors are transmitted to a control center by a measurements for state estimation. Then real power at
communication network. In control center, the data is a bus is increased to introduce error in the system [3].
fed to a state estimator program for further analysis.
State estimation helps to calculate the states (voltage
and bus angles), as well as estimated values of the
measurements in presence of error. It provides a IV. POWER SYSTEM STATE ESTIMATION
snapshot of the power system from available data.
Modeling Of Weighted Least Square Technique
State Estimation is a real time mathematical
methodology or procedure to process real time The electric power transmission system uses
measurements. By using power system model and line wattmeter, VAR-meters, voltmeters and current
flows and power injections and voltage measurement meters to measure real power, reactive power, voltages
values, it is possible to compute an optimal estimate of and currents.
the state vector.
These continuous quantities are monitored by current
and potential transformers on the transmission lines
III. LITERAURE SURVEY and on the transformers and buses of power plants and
substations.
Weighted Least Squares (WLS) and Iteratively
Reweighted Least Squares (IRLS) algorithms are The analog quantities pass through transducers and
implemented on a Ward Hale 6 bus [1] test system. analog-to-digital converters and the digital outputs are
The measurements are obtained by performing power then telemetered to the energy control center over
flow and introducing a random error in the various communication links.
measurements. The data acquired always contains inaccuracies which
are unavoidable since physical measurements cannot
The implementation algorithms is shown in a brief be entirely free of random errors or noise.
way in figure 2.
1)Measurements have certain Error associated with
the true value of measured quantity

[𝒁] = [𝒁𝒕𝒓𝒖𝒆 ] + [𝒆] (1)

[𝒁] = [ 𝒉(𝒙)[𝑥]] + [𝒆] (2)

 5)We select the best estimates of the state variables as
those values which cause objective function f to take
𝒁𝟏 𝒉𝟏(𝒙𝟏, 𝒙𝟐, 𝒙𝟑, … , 𝒙𝒏) 𝒆𝟏 on its minimum value.
𝒁𝟐 𝒉𝟐(𝒙𝟏, 𝒙𝟐, 𝒙𝟑, … , 𝒙𝒏) 𝒆𝟐
. . . 6) For minimizing f , the estimates 𝑥̂ are those values
𝒁= = + (3)
. . . of 𝑥 , which satisfy the equation ,
. . .
[𝒁𝒎] [𝒉𝒎(𝒙𝟏, 𝒙𝟐, 𝒙𝟑, … 𝒙𝒏))] [𝒆𝒎] 𝝏 (𝒆̂ )
[ ] [𝑾][𝒆̂] = 𝟎 (𝟕)
Where, 𝝏𝒙

𝒁 is measurement vector
𝒁𝒕𝒓𝒖𝒆 is true value of measured quantity 7)Substituting 𝒆̂ = 𝒁 − 𝑯 𝒙
̂ in equation (7)
𝒉(𝒙) is the non linear function expressing
̂)
𝝏 (𝒁−𝑯𝒙
measurements in terms of State [ ] [𝑾][𝒁 − 𝑯𝒙
̂] = 𝟎 (8)
𝝏𝒙
𝒙 is the system state vector
𝑒 is the measurement error
𝒎 is the number of measurements 8)Solving for 𝑥̂
𝒏 is the number of state variables

2)Error between actual measurements Z and the true ̂ = (𝒈𝑻 𝑾 𝒈)−𝟏

𝚫𝒙 𝐠𝑻 𝑾 𝒁 (9)
(unknown) value of measured quantities:
The iterative approach is used to solve for the state
[𝒆] = [𝒁] − { [𝑯][𝒙] } = [𝒁] − [𝒁_𝒕𝒓𝒖𝒆 ] (4)
𝒙𝒌+𝟏 = 𝒙𝒌 + 𝚫 𝒙
̂ (10)

3)True values of state cannot be determine so we

calculate estimates of states: Where,
𝜕ℎ
g is Jacobian Matrix,
𝒆̂ = [𝒁] − [𝑯] ∗ [𝒙
̂] (5) 𝜕𝑥
h is Measurement Function
̂ is the estimated value of state variables
𝒙 x are the state variables
k is iteration index
𝒆̂ is the difference between actual measurements and
gT W g is gain matrix
their estimated values
𝒙𝒌 is value of State vector at kth itteration
4)Because of the errors or noise the true values of
physical quantities are never known and we have to 𝒙𝒌+𝟏 is the value of State vector
calculate the best possible estimates of the unknown
quantities. ̂ is update vector of the state vector
𝚫𝒙

𝜕𝑉𝑚𝑎𝑔
We must decide a criterion for calculating the estimates 0
𝜕𝑉
𝑥̂ from which error 𝑒̂ and 𝑍̂ are to be computed. If 𝜕𝑃𝑖𝑛𝑗 𝜕𝑃𝑖𝑛𝑗
algebraic sum of the errors is minimized, then positive 𝜕𝛿 𝜕𝑉
and negative errors would offset one another. So Sum 𝐠= 𝜕𝑄𝑖𝑛𝑗 𝜕𝜕𝑄𝑖𝑛𝑗 (11)
𝜕𝛿 𝜕𝑉
of Squares of Error is minimized. 𝜕𝑃𝑖𝑗 𝜕𝑃𝑖𝑗
𝜕𝛿 𝜕𝑉
𝜕𝑄𝑖𝑗 𝜕𝑄𝑖𝑗
To ensure that measurements from meters of known [ 𝜕𝛿 𝜕𝑉 ]
greater accuracy are treated more favorably than less
accurate measurements, each term in the sum in squares
is multiplied by an appropriate weighting factor W to Where,
give the objective function: Vmag is the voltage magnitude.
Pinj and Qinj are the real and reactive power
𝒇 = ∑𝒏𝒊=𝟏 𝒘𝒊 𝒆𝟐𝒊 (6)
injections respectively.
Pij and Qij are the real and reactive power flow
from bus i to j respectively.
 V. CONCLUSION [13] A. Abur and A. G. Exposito, “Power System State
The main focus of any power utility is the reliable and Estimation: Theory and Implementation”. New York:
Mercel Dekker, 2004.
secure operation of its power system, so it can deliver [14] Mariesa Crow, “Computational Methods for Electric Power
an uninterrupted power supply to its customers. A System”,(Power Engineering Series, vol. 9), CRC press.
reliable estimate of any power system is essential for 2003.
its smooth operation. The technique that estimates the
state of any power system is state estimation.
Estimated state variables are then used in estimating
the line power flows which are then used in system
control centers in the implementation of the security-
constrained dispatch and control of the power system.

Calculated Estimated Values are near to true value as

error associated with estimated values reduces. WLS
State estimation technique calculates accurate value of
State and Measured Quantities with high confidence
despite Measurements are missing, inaccurate or
corrupted by Noise.

VI. REFERENCES

[1] Srinath Kamireddy, Student Member, IEEE, Noel N. Schulz,

Senior Member, IEEE and Anurag K. Srivastava, Member,
IEEE, Comparison of State Estimation Algorithms for
Extreme Contingencies 2008 IEEE
[2] Roque Filipe Mesquita Brandlo (I’ Jose Antonio Beleza
Carvalho ‘I’ Isabel Maria Ferreira state estimation in
transmission line systems
[3] Jiaxiong Chen Yuan Liao and Bei Gou, Study of WLS State
Estimation Convergence Characteristics under Topology
Errors 1013 IEEE
[4] Fang Chen, Xueshan Han,Zhiyuan Pan,Li Han, State
Estimation Model and Algorithm Including PMU
[5] Shivakumar N. R., and Amit Jain, A Review of Power
System Dynamic State Estimation Techniques
[6] Fred C. Schweppe and J.Wildes, “Power system static-state
estimation part I: exact model”, IEEE Transactions on
Power Apparatus and Systems, Vol.PAS-89, pp.120-125,
January 1970
[7] Robert E. Larson, William F. Tinney, Laszlo P. Hajdu and
Dean S. Piercy “State estimation in power systems, part II:
implementation and applications”, IEEE Transactions on
Power Apparatus and Systems vol.PAS-89, pp.355-363,
March 1970
[8] Isabel M.Ferreira and F.P.Maciel,An online Recursive
Algorithm for Power System State Estimation,1991 IEEE
[9] Antonio Exposito and Ali Abur,On the use of PMUs in
Power System State estimation
[10] Carlos Gonzalez-Perez and Bruce F. Wollenberg, “Analysis
of massive measurement loss in large-scale power system
state estimation”, IEEE Transactions on Power Systems,
vol.16, pp.825-832, November 2001.
[11] Do Couto Filho, M.B.; Leite da Silva, A.M. e Falcfio, D.M.,
- Bibliogr.aph,v On Power Svsrem Stale Estitnaiion11968-
1989). IEEE Trans. on PWRS, Vo1.5, no 3, August 1990A
[12] Bose, K.E. Clcments - Red Time Modelling of Power
hierworks. Proceedings of the IEEE,Vol. 75,n”l2. December
1987.

View publication stats