JP2735317B2 - Elevator group management device - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION [Industrial Application Field] The present invention relates to an elevator group management device.
The present invention relates to an apparatus for determining a control method by incorporating a customer request.

[Conventional technology]

Japanese Patent Application Laid-Open No. 59-48369 discloses an apparatus for determining the specification and control method of a conventional elevator group management.
Same as the operation control program as described in
In some cases, service performance is obtained by operating a plurality of elevators in a manner close to actuality using a group management program having similar functions. Japanese Patent Application Laid-Open No. 61-71 discloses an apparatus used for determining the specifications of an elevator installed in a building.
As in No. 82, there is a simulation test device for performing simulation by computer.In addition, as in JP-A-61-263578, an optimal operation analysis of an elevator is performed using a computer to determine the service limit. There is something. The purpose of these conventional examples is to determine the specifications of an elevator that makes the waiting time and long waiting time a certain value or less, and according to requirements, a plurality of targets such as riding time, car congestion degree, etc. There is no idea when deciding the control method.

[Problems to be solved by the invention]

The main purpose of the above prior art is to evaluate the performance when each control method is used, and to determine a control method that can control a plurality of control targets in a well-balanced manner according to a customer request. I haven't been.

An object of the present invention is to provide an elevator group management device capable of accepting a customer request and controlling a plurality of control targets in a well-balanced manner.

[Means for solving the problem]

In order to achieve the above objectives, evaluation indices for evaluating each control method are based on parameters created based on customer requirements, building specifications, usage conditions, elevator specifications, usage conditions. Expressed as a function with parameters obtained by fuzzy inference, it is possible to create an evaluation index that takes into account building specifications, usage, elevator specifications, and usage.

Further, a prediction achievement value is obtained by an inference unit and a simulator which are assumed to be combined with a simulator, and evaluated by the above evaluation index, and is matched to a customer's sense that changes depending on the use status of a building or an elevator. It is possible to determine the group management control method.

[Operation] The sentiment input unit is divided into a pair comparison input and a fuzzy inference unit, and the control targets are compared one-to-one by the pair comparison input, and a pair comparison matrix is created from the comparison result. From this paired comparison matrix, the importance of each control target considered by the customer is obtained by the unique vector method or the like. The fuzzy inference unit obtains a conversion coefficient for each control target from the use and specification of the building, the specification of the elevator, and the usage status. This conversion coefficient indicates, for example, how many seconds of the waiting time correspond to one second of the riding time, and how many seconds of the waiting time corresponds to 1% of the congestion degree in the car. A weighting coefficient of an evaluation index for evaluating the control method is created using the importance and the conversion coefficient. For this reason, when evaluating the control method, the best control can be selected in consideration of both the customer's requirements and the specifications and usage of the building and the elevator.

The control method determining unit includes a part for determining and evaluating the control method, a simulation unit, an inference unit, an environment / traffic flow database, and a control method determination knowledge base. The inference unit uses the environment / traffic flow database and contents, the customer request table created by the sensitivity input unit, and the control method that can satisfy the customer request table using the knowledge in the control method determination knowledge base. A number of candidates are inferred, a simulation is performed in the simulation section using the candidates, a simulation result is evaluated using the evaluation index created in the kansei input section, and a control method table containing the best control method is group-controlled. Send to device.

As described above, the simulation result is evaluated using the evaluation index determined by the sensitivity input unit, and the best control method is selected, so that the difference in the effectiveness of the control method due to the difference in detailed specifications can be reflected. Thus, the effect can be confirmed, so that a control method that can be satisfied by the customer can be determined.

〔Example〕

An embodiment of the present invention will be described below with reference to the drawings. First
FIG. 1 is a software configuration diagram of an embodiment of the present invention.
The personalization support device 1 includes a sensitivity input unit 6, a building / elevator specification input program 7, a control method determination unit 8, and a customer request table 9, and further includes a pair comparison input unit 61, a fuzzy inference unit. 62 、 Fuzzy rule base
It is composed of 63. The control method determination unit 8 includes a control method determination program 81, an inference unit 82, a candidate determination program 83, a simulation unit 84, an environment / traffic flow database 85, and a control method determination knowledge base 86. Individualization support device 1
Determines the control method from the request A of the customer, such as a user or a building manager, and the specification and use state B of the building or elevator, and sends the control method to the group management control device 2 in the form of a control method table 10. Pass. The group management control device 2 determines an assigned elevator in response to a hall call issued from the hall call button 4 according to the passed control method, and issues a command to the car control device 3 to control the group of elevators. Perform control.

The fuzzy inference unit 62 calculates a conversion coefficient between each control target based on the importance input by the pair comparison input unit 61 and the specifications and use status of the building and the elevator input by using the building and elevator specification input program 7. Ask for.

Next, the configuration of the fuzzy inference unit 62 and its inference method will be described with reference to FIGS. 2, 3, and 4. FIG.

FIG. 2 is a diagram showing the configuration of the fuzzy inference unit 62.
The fuzzy inference unit 62 includes a fuzzy inference database 621.
It consists of an associated member function file 622, a fuzzy rule base 623, and an associated member function file 624, a fuzzy inference engine 625, and a fuzzy conclusion file 626.

First, when the specifications and usage status are input by the building and elevator specification input program 7, the corresponding rule group in the fuzzy rule base 63 is read into the fuzzy rule base 623 in the fuzzy inference unit 62, and the external fuzzy rule base is read. The membership function associated with the rule group in 63 is stored in a rule membership function file 624. Next, in the fuzzy inference database 621, the necessary specifications and usage conditions from the building / elevator specification input program 7 and the input result of the pair comparison input 61 are registered as facts. The fuzzy inference engine 625 checks the facts in the fuzzy inference database 621 against the premise of the rules in the fuzzy rule base 623, and if a match is found, the conclusion is renewed as a fact.
Add to 621. If the collated rule is a fuzzy rule having a membership function, the degree of conformity is determined from the actual operation amount and the membership function of the premise. Next, the membership function associated with the conclusion is cut at the degree of conformity, and this membership function and the conclusion are stored in the fuzzy inference database 621 and the membership function file 622 associated therewith. First, when all of the facts registered in the inference database 621 are collated, as shown in FIG. 3, the membership function of the rule having the same subject among the rules having the membership function in the fuzzy inference database 621 is changed. By combining and obtaining the center of gravity, a conversion coefficient, that is, an operation amount is obtained, and then the operation amount and the subject are registered in the fuzzy inference database 621. If the inference result is the final conclusion (in this case, the conversion coefficient), it is also registered in the fuzzy conclusion file 626.

The conversion coefficient is obtained as follows. That is, the center of gravity of the area (∫f (x) dy) indicated by the membership function f (x) obtained by synthesis indicated by the dashed line in FIG. 3 (e) is obtained, and the abscissa coordinate indicating the position of the center of gravity is calculated. Conversion factor.

By repeating the above processing, multi-stage fuzzy inference can be performed by directly treating the membership function, and inference can be performed by treating the fuzzy rules and general rules in the same manner.

FIGS. 3 and 4 show different examples of obtaining the conversion coefficient. FIG. 3 shows an example having a membership function in the premise and FIG. 4 does not.

As described above, for example, the conversion coefficient β _i indicating how many seconds of the waiting time corresponds to what percentage of the boarding time and the car congestion degree is calculated. The sensitive input unit 6, and writes the customer request table 9 summarizes the importance is the input result of the inference result is a conversion factor beta _i and a pair comparison input 61 of Fuaji inference unit 62. The control method determining unit 8 receives the customer request table 9 from the sensitivity input unit 6, determines a control method and the like by inference and simulation, and creates a control method table 10. Fig. 5 shows the fuzzy inference database 621,
The figure shows an example of the fuzzy rule base 623, and FIG. 7 shows an example of the fuzzy conclusion file 626.

Therefore, the control method determining unit 8 will be described with reference to FIGS.

First, the building / elevator specification input by the building / elevator specification input program 7 is stored in the environment / traffic flow database 85. Control method determination program 81
Reads the rules used for inference from the control method determination rule base 86 into the rule base 82 of the inference unit based on the contents of the environment / traffic flow database 85. Next, the building / elevator specifications and the contents of the customer request table 9 are read into an inference database 821 in the inference unit.
Facts in inference database 821 and inference rule base 8
Match the premise (lhs) of the rule in 22.
The conclusion part (rhs) is registered in the inference database 821. At this time, if the conclusion part of the rule has a TAG as an identifier and the TAG1 indicates that the conclusion is used for simulation (OUT or SPC), it is also registered in the conclusion file 824. When all the facts (including added ones) have been matched, the inference ends. When the inference is completed, the simulator interface program 825 looks at the contents of the inference result file 824 and creates a simulation control method table 826. At this time, if the content of TAG2 is determined in advance as a candidate, the control method to be tried is increased by one. Add one. If the content of TAG2 is an instruction, the corresponding portions of all the simulation control method tables 826 have the same conclusion content. The simulation section 84 performs simulation according to the simulation control method table 826, compares and selects the result with the candidate determination program 83, and writes the best control method into the control method table 10.

The best control method is determined using the following weighted norm equation.

lp = Σa _i β _i | f _i − _i | In the above equation, a _i is the importance to the control target values such as the waiting time, congestion degree, and riding time determined by the customer, and β _i is the method already described in FIG. The conversion coefficient f _i is a predicted achievement value for each control target as a simulation result, and _i is each control target value.

Using the prediction achievement values f _i corresponding to the plurality of control methods written in the simulation control method table 826, the weighted norm equation is used to determine the control method with the smallest norm value lp as the best control method. Used for group management control.

In the present invention, since the product of the conversion coefficient β _i and the importance a _i is used as the norm weight, the conversion coefficient β _i itself contains information on the use status of buildings and elevators. The control target value _i may not include information on the use status of the building or the elevator. Conventionally, each of the control target values _i includes information on the use status of a building or an elevator, and a number of control target values are prepared according to the expert knowledge of an elevator designer to meet customer requirements. Because it was impossible to keep it, a rough value was prepared, and the group management control was rough.

In the present invention, even if each control target value is a rough value, the conversion coefficient β _i is obtained by fuzzy inference, which is a fine-grained one according to the use situation of a building or an elevator. Thus, a well-balanced group management control method can be determined in response to customer requests.

FIG. 9 is an example of the inference database 821, and FIG.
The figure shows an example of an inference rule base 822, and FIG. 11 shows an example of a conclusion file 824.

FIG. 12 is an example of an input screen for a pair comparison, and FIG. 13 is an example of a display of a simulation result.

FIG. 14 shows an embodiment in which a pair comparison input and a simulation result are displayed on the same screen. According to this embodiment, although the degree of freedom of the input is limited, the input can be corrected while observing the result, so that the input is facilitated.

FIG. 15 is an example of a customer request table, FIG. 16 is an example showing a part of a building or elevator specification, and FIG. 17 is an example of a control method table.

According to the above embodiment, the conversion factor according to the usage status of the building / elevator is determined by the fuzzy inference unit that can treat the fuzzy rule and the general rule in the same manner,
Since the customer request can be quantified more in accordance with the actual situation, a feasible customer request can be quickly created. In addition, the control method is inferred by the inference unit having two identifiers indicating how to handle the inference result, and the inference and the simulation are combined by creating data used for the simulation by the simulator interface program. A control method determination method can be easily realized. Further, since the use of the rule used for inference in the simulation can be clearly indicated by the identifier, knowledge input and maintenance are easy. Further, the control method table can be easily created by determining the contents of the identifier.

In addition, ambiguous customer requirements are expressed as a function or product of the conversion factor determined from the use of a building and an elevator and the result of a paired comparison input. it can.

〔The invention's effect〕

As described above, according to the present invention, it is possible to obtain an elevator group management device capable of receiving a customer request and controlling a plurality of control targets in a well-balanced manner.

[Brief description of the drawings]

FIG. 1 is a software configuration diagram of the present invention, FIG. 2 is a configuration diagram of a fuzzy inference unit, FIGS. 3 and 4 are diagrams showing a method of obtaining a conversion coefficient, and FIG. FIG. 6 is a diagram showing an example of a fuzzy rule base, FIG. 7 is a diagram showing an example of a fuzzy conclusion file, FIG. 8 is a configuration diagram of an inference unit, Fig. 9 shows an example of an inference database, Fig. 10 shows an example of an inference rule base, Fig. 11 shows an example of a conclusion file, and Fig. 12 shows a paired comparison. FIG. 13 is a view showing an example of an input screen, FIG. 13 is a view showing an example of display of a simulation result, FIG. 14 is a view showing an example of an input / output screen, FIG.
FIG. 16 is a diagram showing an example of a customer request table, FIG. 16 is a diagram showing an example of a part of a building / elevator specification, and FIG. 17 is a diagram showing an example of a control method table. 1 ... individualization support device, 2 ... group management control device, 3 ...
Unit control device, 4 ... hall call button, 6 ... sensibility input unit, 7 ... building / elevator specification input program,
8: Control method determination unit, C: User interface program.

 ──────────────────────────────────────────────────続 き Continuing on the front page (72) Inventor Kiyoshi Nakamura 4026 Kuji-cho, Hitachi City, Ibaraki Prefecture Within Hitachi, Ltd.Hitachi Research Laboratory Co., Ltd. Inside the plant (72) Inventor Takaaki Uejima 1070 Ma, Katsuta-shi, Ibaraki Pref. Inside the Mito Plant, Hitachi, Ltd. (56) References JP-A-1-231776 (JP, A) JP-A-2-305774 (JP, A)

Claims (4)

(57) [Claims] In an elevator group management control device for managing a plurality of elevators, a means for taking in specifications and usage of buildings and elevators, a request from a customer is accepted, and an input result is input to specifications and usage of buildings and elevators. Means for digitizing using a membership function relating to, means for determining a control method for realizing the request using the digitized result, and means for group management control of each elevator according to the control method obtained by the determining means. A group management device for an elevator, comprising: 2. The means for accepting a request from a customer inputs the relative importance of a plurality of control targets by comparing them one-to-one, and obtains the importance of each control target from the input result. A conversion coefficient between each control target is obtained from a fuzzy rule having a membership function indicating a specification of a building / elevator, a use state and a feeling of a user or a building manager regarding each control target, and the conversion coefficient and importance are used. 2. The elevator group management device according to claim 1, wherein the customer request is quantified. 3. The means for accepting the customer's request includes: an inference database having a fact to which a membership function is added, a rule base having a rule to which a membership function is added, the rule base and the inference database. 2. The elevator group management device according to claim 1, wherein the conversion coefficient is obtained by an inference engine that makes inferences using the group. 4. An elevator group management apparatus for managing a plurality of elevators, the difference between a control target value corresponding to a customer request and a predicted achievement value obtained by simulating whether or not the control target can be achieved. A control method that minimizes the sum of a plurality of control targets of the importance of the value of the control target and a conversion coefficient obtained by quantifying the control target using a membership function regarding the specification of a building and the use status of an elevator. An elevator group management device comprising means for controlling each elevator and realizing the customer request.