DE69419891T2 - Intelligently distributed control for lifts - Google Patents

Abstract

Distributed processing units (DPU) are applied to elevator dispatching and back-up architecture. They are localizing the processing functions by sensing and controlling local devices and serially transmitting up-dates to other DPUs via one or more communication media (18). This makes it possible to confine the wiring to local levels eliminating the need for a centralized processor. The distributed processing units (DPU) are designed as floor processing units (FPUa,FPUb,...) one on each floor (E1...), car processing units (CPUa,CPUb,...) in each car (2), group processing units (GPUa,GPUb,...) and signalling processing units (SPUa,SPUb,...) in the machine room (22). They all possess intelligence to communicate and to perform local control algorithms. The distributed control is structured as a neural network, the nodes being implemented by the distributed processing units (DPU). Depending on the function to be processed, there are main nodes (A0,B0,C0...) with additional back-up nodes (Aa,Ba,Ca...) and auxiliary nodes (A1,A2,A3...). The inventive, distributed control is characterized by modularity, configurability and simplicity, providing reduced development cycle, ease of maintenance and increased reliability. <IMAGE>

Description

The present invention relates to a new and improved intelligent distributed control for elevators comprising: a plurality of elevators arranged in a conventional manner for serving a plurality of floors of a building, a group control having floor call registering means arranged on each of the floors for sending floor calls for ascent and descent on each of the floors for exchanging signals with each of the elevators and for controlling the operation of the elevators in response to the floor calls and in response to signals received from the elevators, a car for each of the elevators having car call registering means for service requested by passengers therein, a car drive for providing and stopping movement of the car and a car control for providing signals indicative of the state of the car for controlling the movement of the car, means for causing the car to move in a selected upward or downward direction and for controlling the car in response to the state of the car and to stop on signals received from the controller, which

the group control comprises a signal processor which reacts to signals indicating the status of each of the cars in order to provide a rating calculation for each car when a floor call is generated, and on the basis of the result of the rating calculation an optimal elevator car is selected and dispatched to answer the floor call. Such a control is known from US-A-5,012,899.

An elevator control system is traditionally a centralized system where the operations are controlled by an intelligent station in the system. This station can be located in the machine room, inside or on the car, etc.

Such a control system is known from the company's own US patent No. 5,305,198, issued on April 19, 1994 (IP 389). In this system, destination calls are definitively and immediately assigned to the individual elevators for servicing the call in accordance with higher-level and lower-level functional requirements. These assignments are immediately displayed on the call input floors. A weighted sum corresponding to higher-level functional requirements is formed from partial service costs and modified with the help of variable bonus and penalty point factors into service costs in terms of lower-level functional requirements, and a destination call is assigned to the elevator with the lowest service costs. This method is carried out on an industrial computer using a destination call assignment algorithm with lower-level algorithms for bonus and penalty point tracking and the Cost calculations are implemented. This calculation of the operating costs takes place in the cost calculation algorithm according to a special cost formula, with the adjusted bonus and penalty point factors having a multiplicative effect on a six-term partial cost sum. The state of the art uses a computer as a centralized control for a lift group consisting of three lifts A, B, C and is not operated with car calls, but exclusively with destination calls. The various elements of the lift group are connected to the controlling computer via a bus system: the measuring and setting elements via a lift bus, the elements in each car via a car bus and the numeric keypads on the floors for entering destination calls via a floor bus. The lift bus, the car bus and the floor bus form a triple bus system that is connected to the computer via a special interface. The centralized computer can have any suitable configuration, such as a separate computer for each lift car, with one of the computers also controlling the group functions, a single computer for the entire lift group or a single computer for two or more lift groups. All of these computer configurations have more or less centralized control in common.

A major disadvantage of this type of system configuration is that, despite the use of a bus system, a significant amount of wiring is required. Wiring must be done from remote locations in the system to a centralized location. This process is not only expensive, it can also be error-prone and laborious. Another disadvantage of such a system is that it requires considerable effort in terms of software development, testing and maintenance. When the elevator system is enlarged, that is, when the number of floors and cars is increased, the centralized computer and other elevator control devices may be overloaded. In this case, the load is not balanced and the processing efficiency of the computer for the entire system is poor. With a single centralized elevator control station, there is no way to distribute and average the load of control functions and data processing.

Accordingly, the object of the present invention is to overcome the above-mentioned disadvantages by creating a completely modular system with greatly reduced overall wiring effort.

In addition, the control system according to the invention should be structured in such a way that the system has increased efficiency and reliability for all functions relating to set operation. This problem is solved according to the invention by the means as characterized in the version of the independent claim. Advantageous developments are indicated in the dependent claims.

The problems and deficiencies of the prior art centralized controllers are solved according to the present invention by the use of distributed processing units (DPUs), which additionally offer the following advantages: A first advantage can be seen in the modularity and configurability of a distributed controller. A DPU-based elevator controller localizes the processing according to the functions. This creates a controller in which the processing for a function is done locally, in the respective part of the control system. With generally defined interfaces, different functions can be added just by adding the necessary hardware and DPUs. This makes it easier to build a system with the desired features without affecting other DPUs and thus the modularity and configurability. This also makes it easier to add more functions to a given order in the future with a minimum of engineering work. A further advantage can be seen in the simplification of the control structure resulting from distributed processing. The control system is divided into various modules, each of which performs a predetermined set of tasks. The development and maintenance of the control system is thus simplified due to its modular nature. By specifying predetermined interfaces, each module can be developed independently and simultaneously to reduce the overall development time of the system. This reduces time to market and development costs and simplifies software maintenance. The system according to the invention has also been found to have improved overall reliability and safety since there is no single point of failure and degraded modes of operation are possible.

The above and other advantages of the present invention will become apparent to those skilled in the art from the following detailed description of a preferred embodiment with the aid of the accompanying drawings. They show:

Fig. 1 is a schematic diagram of a conventional elevator group, in which an industrial computer is used to perform centralized control of three elevators.

Fig. 2 is a schematic representation of an elevator system in which a distributed control according to the present invention based on a neural network is used.

Fig. 3 is a schematic representation of the neural network structure used to implement the distributed control according to the invention. In Fig. 1, the elevators of an elevator group are designated A, B and C, with a car 2 being guided in an elevator shaft 1 for each elevator and driven in a known manner by an elevator motor 3 by means of an elevator cable 4 to serve sixteen floors E1 to E16. Each drive 3 is controlled by a drive controller, whereby the target value generation, the regulation functions and the start/stop initiation are all implemented with the aid of an industrial computer 5. Measuring and setting elements 6 are connected to the industrial computer 5 via a first interface IF1 and an elevator bus 7. Each car contains a load measuring device 8 to determine when passengers enter or leave the elevator car, a call display device 9 which reports the current operating status Z of the car, a stop display 10 and a car control panel 11. The devices 8, 9, 10 and 11 are connected to the computer 5 via a car bus 12.

Car calls are recorded in the elevator cars A, B and C by means of suitable push-button arrangements in the car control panel. They are then serialized and transmitted together with other information concerning the car via the car bus 12 and the CIF interface to the industrial computer 5.

On the floors E1 to E16, call registration devices 8 are provided in the form of suitable push buttons 13, such as a floor call push button 14 "Up" arranged on the lowest floor E1, a floor call push button 15 "Down" arranged on the highest floor E16, and floor push buttons 16 "Up" and "Down" arranged on each of the intermediate floors E2 to E15. The floor calls are serialized like the car calls and transmitted via the floor bus 17 and the input interface ICF to the industrial computer 5, where they are assigned to the individual cars 2 in the sense of a demand function profile using a special floor call assignment algorithm for operation.

Fig. 2 shows the elevator control system based on distributed control concepts. It uses Neuron Chip-based Lonworks (local operating networks) from Echelon Corporation, which is a new technology that promises better opportunities to create distributed control systems at lower cost using unique means of implementing distributed control algorithms. The underlying principle of the technology is to create a system with Distributed Processing Units (DPUs). These DPUs have intelligence for detecting and controlling local devices and sending updates to other DPUs in the system. All distributed processing units DPU operate automatically, independently and autonomously under the management of software stored in each of them. They are easily connected to each other via one or more available communication media 18 and thus have a common message-based communication protocol. This makes it possible to limit wiring to local levels and to transmit device status serially to different points within the system. This greatly reduces the overall wiring effort. Each DPU of such a system has intelligence not only for transmitting messages between different points within the system but also for executing control algorithms. By properly distributing control functions to different DPUs, a centralized processor is no longer needed.

This is illustrated in the circuit diagram of Fig. 2. The control system is here composed of several intelligent DPUs which are interconnected via a communication medium 18. The number of DPUs varies according to the requirements of the control system, which need not be based on the number of elevators and floors, but on the required processing capacity. The functions performed by any given DPU, however, do not change. On three floors E1, E2, E3, floor processing units FPU are arranged in the corridor installation boxes 20 of each floor E1, E2, E3 and in the fire control center 21. In the elevator cars 2, car processing units CPU for door operations, position indication, landing system and car call features are arranged. At the machine room 22, further group processing units GPU and signaling processing units SPU are arranged. The floor processing units FPU and the car processing units CPU carry out functions associated with their sensor devices and their actuator devices and the necessary control functions. In addition, they broadcast the latest states of the critical devices to inform other DPUs in the system. The information thus obtained from other DPUs is used together with their local device states to make control decisions for which it is responsible. The group processing units GPU mainly contain floor call allocation functions, whereas the signalling processing units SPU carry out transmission functions for Interprocess communication. In this way, a fully modular system with distributed control functions can be created.

An example of how a typical elevator operation is carried out by such a system is described with reference to Fig. 2. When a corridor call button 25 is pressed by a passenger on the third floor, this is detected and recorded by the floor processing unit FPUa arranged in the corridor installation box 20 on that floor if an elevator A, B, C is in operation. This information is then transmitted to a first group processing unit GPUa in the machine room 22. The group processing unit GPUa is responsible for carrying out the rating calculation for determining floor call assignments to cars 2 and for carrying out the group management of the cars 2 on the basis of the result of the rating calculation by controlling the drive unit 26, which in turn controls the movement of the elevators A, B, C. After receiving the information regarding the demand on the third floor, the group processing unit GPUa initiates commands to move the assigned elevator, which in this case shall be elevator A, towards the third floor. As the elevator A moves through the building, the car processing unit CPUc, which monitors the landing system, updates other distributed processing units DPUs as the elevator A passes the floors E1, ... Upon receiving these updates, the third floor floor processing unit FPUa changes the position indicator 27 to the corresponding floor position and direction. At the same time, the first group processing unit GPUa also receives the landing system update from the car processing unit CPUc and decides on the basis of this information whether to continue the journey or stop at the next floor. If the next floor is the destination floor, the second group processing unit GPUb changes the commands to the drive unit 26 to stop at that floor. After the elevator has started to decelerate, the second group processing unit GPUb broadcasts this information over the network. When the floor processing unit on the third floor FPUa receives this information, it cancels the floor call that has just been answered and at the same time switches on the floor lamp 28 for it. When the elevator reaches exactly the height of the floor, the group processing unit GPUb receives an update from the landing system car processing unit CPUc, thereby instructing the drive unit 26 to stop the elevator there, and the second group processing unit GPUb informs the door processing units CPUa and CPUb that the elevator is on the destination floor. At this point, CPUa and CPUb, which are responsible for controlling the door operation, instruct the doors to open and then close after a predetermined time. This satisfies the demand on the third floor. The typical elevator operation described involves the exchange of data between the various distributed processing units DPUs. For this purpose, signaling processing units SPUa, SPUb, ... are provided to cause the group processing units GPUa and GPUb to communicate with each other through a first data field 29 and to cause the car processing units CPUa, CPUb, ... to communicate with each other through a second data field 30. As illustrated in the above example, a completely modular system can be created without the need for a central processing unit such as that designated 5 in Fig. 1. By distributing the processing responsibilities of the system, tasks become much simpler and the load balance can be averaged. This simplicity is directly reflected in the system development cycle and the ease of maintenance. In practice, for load averaging, a distributed processing unit with a heavy load can be temporarily excluded from process execution. Even if only one of the group processing units GPUa, GPUb, ... is operational, group control can be carried out, thus ensuring reliability in this respect. Through cooperative distributed control realized by the individual distributed processing units DPU, high reliability and high system efficiency are achieved.

Fig. 3 illustrates the neural network structure used to realize the distributed elevator dispatch control and the backup architecture. A multi-car dispatch model was adopted by choosing a neuron chip MC 143150 for low cost and performance. The car to be assigned to a landing call is selected based on the results obtained using the neural network corresponding to the neurons of the human brain. The neural network includes an input layer, an output layer and an intermediate layer provided between the input and output layers. In the intermediate layer, weighting factors are applied when linking signals from the nodes of the input layer and when distributing signals to the output nodes. The weighting factors are variable and are changed and corrected accordingly by learning to obtain a more appropriate car assignment. At different times or under different detected conditions of the Depending on the passenger load, different sets of learning factors can be applied. Learning (correction of the network) can be done using the backpropagation technique. Backpropagation is a technique for correcting the weighting factors using errors between the output data of the network and desired output data, created from monitored data or control objective values. Each main node is designed to contain its own Estimated Time of Arrival software to execute a Token Ring algorithm for car-to-car communication and the backup architecture. Both guarantee the reliability of the set service.

In this dispatch model, the nodes A0, B0, C0, D0, ... are the main nodes (MC143150) for all set cabins, nodes A1, A2, A3, ... are the auxiliary nodes (MC143120) for each cabin, and the nodes Aa, Ba, Ca, Da, ... are the reserve nodes (MC143150) for the main nodes of all cabins A, B, C, D.

Each master node contains its own ETA calculation software to run a token ring algorithm for car-to-car communication to guarantee the reliability of set service, instead of having only one node calculating ETA. Neuron Network Management also ensures the simultaneity of signal transmission among the auxiliary nodes A1, A2, A3, ... to the master nodes A0, B0, C0, ... The 3150 chip provides 64 Kb for a user program and goes over the common network channel 34 to communicate with other nodes.

For the critical nodes such as the main nodes A0, B0, C0, ..., additional reserve nodes Aa, Ba, Ca, ... can be used as shown. Each node pair 35 consists of a main node and a reserve node. It starts for incoming data and is followed by a multiplexer 36 for outgoing data as well. In the reserve node side, the conformance checking software constantly sends out signals to check the common network channel 34 and the application I/O channel 37 of the main node. When a fault is found, a signal is sent out to select the reserve nodes Aa, Ba, Ca... from the node pair 35 for data communication. When the system is installed, all the nodes have their own labels, even the reserve nodes Aa, Ba, Ca, ... All the related nodes can search for reserve nodes when a common fault occurs.

REFERENCE LIST

1) Elevators; elevator group A, B, C

2) Elevator shaft 1

3) Cabin 2

4) Drive 3

5) Elevator cable 4

6) 16 floors E1... E16

7) Industrial computer 5

8) Measuring and adjusting elements 6

9) 1. Interface IF1

10) Elevator bus 7

11) Load measurement equipment 8

12) Operating state Z

13) Operating status signaling equipment 9

14) Stop indicator 10

15) Cabin control panel 11

16) Cabin bus 12

17) CIF interface

18) Pushbuttons 13

19) Floor call push button "Up" 14

20) Floor call push button "Down" 15

21) Floor call push button "Up" and "Down" 16

22) Floor bus 17

23) Distributed Processing Units DPU

24) Communication media 18

25) Floor processing units FPU

26) Corridor installation box 20

27) Fire Control Center 21

28) Cabin processing units CPU

29) Group processing units GPU

30) Signal processing units SPU

31) Engine room 22

32) Sensor device 23

33) Actuator device 24

34) Corridor call button (on the third floor) 25

35) Floor processing unit: Floor 3 FPUa

36) Floor processing unit: Floor 2 FPUb

37) Floor processing unit: Floor 1 FPUc

38) Fire control processing unit FPUd

39) First group processing unit GPUa

40) Second group processing unit GPUb

41) Drive unit 26

42) Front door cabin processing unit CPUa

43) Rear door cabin processing unit CPUb

44) Landing system cabin processing unit CPUc

45) Position display cabin processing unit CPUd

46) Car call feature car processing unit CPUe

47) Position indicator 27

48) Third GPU processing unit GPUd

49) Floor light 28

50) First data field 29

51) Second data field 30

52) Main nodes A0, B0, C0, ...

53) Auxiliary nodes A1, A2, A3, ...

54) Reserve nodes Aa, Ba, Ca, ...

55) Common network channel 34

56) Node pair (main node/reserve node) 35

57) Multiplexers 36

58) Application EIA Channel 37

59)

60)

61)

62)

63)

64)

65)

Claims (6)

1) Intelligent distributed control for elevators (A, B, C), comprising: a plurality of elevators (A, E3, C) arranged in a conventional manner for serving a plurality of floors (E1, E2, ...) of a building, with a group control (GPU) with floor call registering devices (13) arranged on each of the floors for sending floor calls for ascending and descending on each of the floors (E1, E2, ...) for exchanging signals with each of the elevators (A, B, C) and for controlling the operation of the elevators (A, B, C) in response to the floor calls and in response to signals received from the elevators (A, B, C), further comprising a car (2) for each of the elevators (A, B, C), a car drive (3) for providing and stopping the movement of the car (2) and a car control for providing the Signals indicative of the state of the car (2) for controlling the movement of the car, means for causing the car (2) to move in a selected upward or downward direction, and for stopping the car (2) in response to signals indicative of the state of the car (2) and to signals received from the controller, whereby the group controller (GPU) comprises a signal processor (SPU) responsive to signals indicative of the state of each of the cars (2) to provide a rating calculation for each car (2) when a floor call is generated, and on the basis of the result of the rating calculation an optimal elevator car (2) is selected and dispatched to answer the floor call, characterized - in that the control comprises distributed processing units (DPU) and is preferably constructed as a neural network, with input, intermediate and output layers, whereby the nodes of the neural network are implemented by the distributed processing units (DPU), - by floor processing units (FPUa, FPUb, ...) arranged on the floors (E1, E2, ...) for controlling the corridor installation boxes (20) of each floor (E1, E2, ...) and autonomous input/output of information associated with the floor (E1, E2, ...), - by car processing units (CPUa, CPUb, ...) arranged in each car (2) of the lifts (A, B, ...) for controlling each car (2) and autonomously inputting/outputting information associated with this car (2), - by group processing units (GPUa, GPUb, ...) arranged in the machine room (22) for carrying out the evaluation calculation for determining the floor call assignments to cars (2) and for carrying out the group management of the cars (2) on the basis of a result of the evaluation calculation; and - by one or more signal processing units (SPUa, SPUb, ...) provided to cause the group processing units (GPUa, GPUb, ...) and the floor processing units (FPUa, FPUb, ...) to communicate with each other. 2) Intelligent distributed control according to claim 1, characterized in that that in each cabin (2) a first door processing unit (CPUa) for operating the front door, a second door processing unit (CPUb) for operating the rear door, a landing system processing unit (CPUc) for the landing system, a position processing unit (CPUd) for position display and a cabin call processing unit (CPUe) for cabin call features are provided. 3) Intelligent distributed control according to claim 1, characterized in that the device status is transmitted serially to the various distributed processing units (DPU) within the system. 4) Intelligent distributed control according to claim 1, characterized in that the intermediate layer contains first weighting factors between the individual nodes of the input layer and the individual nodes of the intermediate layer and second weighting factors between the individual nodes of the intermediate layer and the individual nodes of the output layer. 5) Intelligent distributed control according to claim 1, characterized in that the neural network contains main nodes (A0, B0, C0, ...) and auxiliary nodes (A1, A2, A3, ...), whereby each main node (A0, B0, C0, ...) is connected in parallel to a reserve node (Aa, Ba, Ca, ...). 6) Intelligent distributed control according to claim 1, characterized in that each main node (A0, B0, C0, ...) contains its own ETA software to execute a token ring algorithm for cabin-to-cabin communication.