JP3215426B2 - Elevator group management system - Google Patents

Description

Description: TECHNICAL FIELD The present invention relates to an elevator group management system, and more particularly to an apparatus for efficiently searching for an optimal combination of control parameter values.

BACKGROUND ART An elevator group management system is a system for efficiently operating a plurality of elevators according to various traffic conditions in a building. The group management device in this system
Operation control such as elevator assignment is performed according to the group management algorithm. Here, the group management algorithm executes and controls all functions and operations related to elevator operation, including elevator assignment control.

The group management algorithm includes various control parameters, and in order to achieve efficient driving, appropriate numerical values are set for the parameters according to various traffic conditions in the building. Must be assigned.

In the assignment control of a hall call (a call from an elevator hall), which is one of the basic functions of group management, when a new hall call is registered, the new hall call and the already registered hall call are In order to evaluate a service state such as a waiting time or a forecast, an evaluation value Em is obtained for each elevator (car) according to an allocation evaluation function described later. Then, the elevator having the minimum evaluation value Em is selected as the assigned elevator. Then, a hall lantern or the like provided at the landing is turned on, and a guidance display (this is referred to as a forecast) is provided to the waiting customer before the assigned car arrives.

As a function for calculating the above-mentioned assignment evaluation value Em, for example,
There is the following formula [1]. Here, i is the hall call number, and m is the elevator number.

However, Em: an assigned evaluation value when a new hall call is assigned to elevator m W (i): a predicted waiting time of hall call i when a new hall call is assigned to elevator m M (i): new for elevator m Fullness probability of hall call i when a hall call is assigned (0 ≦ M (i) ≦ 1) Y (i): Probability of a non-forecast of hall call i when a new hall call is assigned to elevator m (0 ≦ Y) (I) ≦ 1, the out-of-forecast means a phenomenon in which an elevator other than the predicted elevator arrives first) Pm: Penalty when a new hall call is assigned to elevator m Bm: Penalty when a new hall call is assigned to elevator m Bonus Ca: fullness evaluation coefficient Cb: forecast outlier evaluation coefficient Here, the fullness evaluation coefficient Ca is the waiting time evaluation value W (i) ²
Is a weighting coefficient of the fullness evaluation value M (i) with respect to.

In addition, the forecast deviation evaluation coefficient Cb is calculated as the waiting time evaluation value W (i)
² is a weighting coefficient of the forecast outlier evaluation value Y (i) for ² , and if the value is increased, allocation can be performed with more emphasis on prevention of forecast miss than in waiting time.

Examples of the priority assignment function using the penalty Pm in the above equation [1] include the following [2] riding time priority assignment function and [3] power saving priority assignment function as described below.

[2] The boarding time priority assignment function is a function that makes it difficult to assign a call from a landing halfway to an elevator having many car calls. For example, a value calculated by (riding time priority Pa) × (call number Nm) is set as the penalty Pm.

[3] The power saving priority assigning function is a function that makes it difficult to assign a new call to a suspended elevator. For example, the value indicated by the power saving priority Pb is set as the penalty Pm for the idle elevator, and 0 is set for the other elevators.

In addition, the priority assignment function using the bonus Bm in the above equation [1] includes, for example, [4] a proximity elevator priority assignment function, [5] a light load elevator priority assignment function, and
[6] There is a specific elevator priority assignment function.

[4] The proximity elevator priority assignment function is a function that facilitates assignment of an elevator (proximity elevator) near the operated button. For example, the value indicated as the proximity elevator priority Ba is set as the bonus Bm for the proximity elevator, and the bonus Bm is set for the other elevators.
Is set to 0.

[5] The light-load elevator priority assignment function is a function that makes it easy to assign an empty elevator or an elevator with a small load (light-load elevator). For example, for a lightly loaded elevator, the value indicated by the lightly loaded elevator priority Bb is set to the bonus Bm, and for other elevators, 0 is set to the bonus Bm.

[6] The specific elevator preferential allocation function is a function that makes it easy to allocate a specific elevator. For example, the value indicated by the specific elevator priority Bc is a bonus for underground elevators, rooftop elevators, observatory elevators, etc.
Bm is set to Bm, and for other elevators, 0 is set to the bonus Bm.

As described above, Ca, Cb, Pa, Pb, Ba, Bb, and Bc are parameters for group management related to the assignment evaluation function [1].

Even when the assignment is performed using the above assignment evaluation function [1], an unexpected call may occur after that and a long wait may occur. Therefore, the group management system has [7] additional assignment function and [8] assignment change function.

[7] The additional allocation function for a long waiting call is a function of performing an additional allocation that causes another elevator that can service earlier than the currently allocated elevator to perform rescue.

[8] The assignment change function for long waiting calls is to shift the assignment (and forecast) of calls in long waiting state exactly to the rescue elevator. A determination reference value DL is set to detect whether or not a long wait has occurred.

Each elevator in the group management system is
[9] It has a packed automatic passage function, and when the car load exceeds the reference value DB, the car automatically passes through the landing even if it has already been assigned. In addition, for calls that are automatically passed because of full capacity, [10] rescue operation is performed by the assignment change function.

[10] The allocation change function for a full-passing call is to transfer the allocation and forecast of a hall call automatically passed when it is full to another rescue elevator. Changing the forecast to a new assigned car is referred to as forecast change.

As described above, the above DL and DB are also parameters for group management.

Also, various control parameters are used in operations other than the operation related to the hall call. For example, various control parameters are used as conditions for selecting or canceling the following operation patterns.

[11] Selection / elimination of driving during commuting Driving during commuting is based on the number of car calls registered for the starting elevator on the main floor (main floor) after the start time of the commuting time zone. Selected when DIUPC or more, while canceled when the end time of the working hours has passed.

[12] Selection / elimination of up-peak operation Up-peak operation is selected when the number of occupants on the main floor exceeds the first criterion value DUP1 and an elevator departs, while riding on the main floor Number of people is the second criterion value DU
If the elevator departing beyond P2 is during the judgment time DUPT,
It is canceled when no one is generated.

[13] Selection and cancellation of down-peak driving Down-peak driving is selected when an elevator descends with passengers of the first criterion value DUP1 or more, while passengers with the second criterion value DDR2 or more are selected. The elevator is canceled when no elevator is lowered during the determination time DDPT.

Each operation pattern includes the following control, but also includes control parameters.

[14] Driving at work ・ Distribute the number of elevators specified as DIUPN on the main floor.

-In the departure adjustment operation of delaying departure with the door open even if a call to be answered to the first floor elevator starts, the departure time (= the time to keep the door open) is set to the reference value DIUPT. Set.

-On the main floor, the elevators designated as the number of door-open standby units, DIUPW, are placed in the door-open standby state, and the other elevators are placed in the door-close standby state.

[15] Up-peak operation ・ Distribute elevators of the number specified as DUPN to the main floor.

[16] Down-peak operation ・ When performing prediction calculation of waiting time, for hall calls in the direction toward the main floor, only the amount corresponding to the priority DDPE is applied.
Calculate the expected wait time longer.

[17] Decentralized standby operation-When an empty elevator occurs, the elevator is set to wait in advance on the floor where the next call is likely to occur to reduce the waiting time, and when the congestion operation is not selected. Selected.

-When the state where there are empty elevators equal to or more than the specified number DOHN continues for the specified time DOHT or more, the distributed standby operation is executed.

The floor at which the elevator is on standby (standby floor), the number of units on standby, and the like may also be used as control parameters.

Furthermore, in the following additional control, a control parameter is included in order to control the number of operating units.

[18] Power saving operation In this operation, the number of operating vehicles is automatically reduced according to the service status to save power. If the average waiting time for the last 5 minutes is less than or equal to the first service reference value DESW1, the current The number of operating vehicles is decreased by one, and when the number of operating vehicles is equal to or more than the second service reference value DESW2, the number of operating vehicles is increased by one.

As described above, the group management algorithm includes many parameters. These parameters are prepared in order to satisfy various control purposes such as shortening of waiting time, improvement of forecast accuracy, improvement of user comfort, and power saving. However, since some parameters have conflicting control objectives, the performance of group management greatly depends on the combination of numerical values assigned to each parameter.

In other words, it is necessary to quickly search for an optimal combination of parameter values in order to perform an efficient group management operation according to various traffic conditions of a building that changes every moment and various requests of a user.

In the following description, a combination (numerical sequence) of parameter values is called a “parameter value set” or simply “set”.

As a conventional optimal set search method, there is a brute force method in which all combinations of parameter values (that is, all sets) are examined (for example, Japanese Patent Publication No. 4-51475, Japanese Patent Application Laid-Open No. 57-57168). No., see).

In the brute force method, there is almost no problem when the types of parameters are small.

However, when the number of types of parameters increases, the number of combinations of parameter values to be considered increases dramatically. Therefore, it is extremely difficult to study all combinations and select an optimal set. The details will be described below.

In the brute force method, the number of parameter types is M, and the number of possible values of the parameter is L. Suppose that M = 3
And L = 6, the total number of simulations is L ^M = 21
6 times. Therefore, when the types of parameters and the number of possible values are large, even if a simulation is performed,
Also, even if a trial run of an elevator is performed with an actual group management device, it takes a very long time to determine an optimal combination of parameter values, which is not practical.

Thus, in the system described in Japanese Patent Publication No. 5-24067, a proposal has been made to reduce the number of simulations. That is, in this system, for example, when there are two parameters, the optimum value of the first parameter is obtained first, and then the optimum value of the second parameter is obtained with the value of the first parameter fixed at the optimum value. That is the method. This is called a sequential method.

According to this sequential method, the total number of simulations when M = 3 and L = 6 is L × M = 18,
The number of simulations is greatly reduced as compared with the brute force method.

However, the sequential method is effective only when there is no correlation between the parameters, and cannot be applied to a case where many strongly correlated parameters are included such as a parameter group for group management control.

DISCLOSURE OF THE INVENTION The present invention has been made in view of the above conventional problems, and for a group management parameter group having a special property having a strong correlation, even if the number of parameter value sets is enormous,
The goal is to efficiently search for the optimal set.

In addition, the present invention relates to a “genetic algorithm” (Genetic algorithm).
Based on a general technology called Algorithm), it is intended to realize a unique search method for searching for an optimal set.

(1) In order to achieve this object, a search device according to the present invention comprises: storage means for storing a plurality of sets; one or a plurality of sets selected as a parent from the storage means; Generating means for generating one or a plurality of new sets inherited from the beginning, evaluation means for obtaining an execution result when the group management algorithm is executed using the new set as a group management performance value, and storage means Selecting means for improving a plurality of sets stored in the storage means by using the addition of the new set and the deletion of the inferior set from the storage means in combination; and Extracting means for extracting the optimum set from the grouped sets based on the group management performance value; and a precondition for a search given at the start of the search for the optimum set changes. In particular on the basis, including a re-search determining means for instructing a re-search of the optimum set for the search apparatus.

According to the above configuration, the probability of generating an excellent set can be increased by generating a genetic set and selecting a good set, and at the same time, a child set (new set) that inherits only the good properties of the parent set. Only the data can be stored in the storage means. That is, by repeating a series of cycles, a plurality of sets stored in the storage unit can be sequentially updated to improve the quality. Then, finally, the optimum set is extracted from the storage means based on the group management performance value. Each numerical value constituting the optimal set is substituted into each parameter in the group management algorithm, and group management such as elevator assignment is performed.

As described above, according to the present invention, it is possible to efficiently search for the most excellent set or the set having the content very similar to the best set. That is, the amount of calculation and the number of simulations can be reduced, so that the search can be speeded up.

(2) In the present invention, the generation unit includes: a set between two sets selected from the storage unit;
A numerical value exchange means for generating two new sets by exchanging some numerical values with each other, and replacing some parameter values in one set selected from the storage means with new random numbers generated And a generating method selecting means for stochastically selecting the numerical value exchange and the new value replacing.

According to the above configuration, the generation of the new set is performed by randomly selecting “crossover” by the numerical value exchange unit and “mutation” by the new value replacement unit.

In general, crossover has the property of converging the solution, and mutation has the property of producing diversity in the solution. Therefore, according to the crossover, the contents of the set group stored in the storage means can be converged, but on the other hand, the diversity of the set group is lost at an early stage, resulting in a local solution and a true solution. May lose the solution (optimal set). In that case, according to the mutation, it is possible to depart from the local solution. In that sense, crossover and mutation are complementary.

Mutations, on the other hand, have the property of destroying good solutions sought by crossover. In that sense, crossover and mutation are in competition.

Therefore, it is necessary to appropriately set the ratio between the probability that the crossover is selected (crossover rate) and the probability that the mutation is selected (mutation rate).

In any case, by appropriately using both crossover and mutation, the advantages of both can be fully utilized and the probability of generating an excellent new set can be improved.

However, in some cases, it is also possible to search for the optimal set only by crossover or mutation.

(3) In the present invention, the generation unit includes: a parent selection unit that selects one or two sets from the storage unit; and a parameter in which the numerical value exchange or the new value replacement is performed for the one or two sets. And parameter selection means for selecting.

According to the above configuration, when the crossover is selected, the parent selection unit stores the two parent sets (set pairs) from the storage unit.
Is selected, and if a mutation is selected, one parent set is selected from the storage means. The parameter selection means selects a parameter position (crossover position or mutation position) at which parameter values are replaced in crossover and mutation.

(4) The above-mentioned parent selecting means selects a parent based on parent selection criterion information for increasing the generation probability of an excellent new set. By using the parent selection criterion information, the probability of generating an excellent new set can be increased.

(5) If the distance between sets is used as parent selection criterion information,
Similarity between sets can be used as a basis for selection. This allows
The diversity of the new set to be generated can be prioritized, or the convergence of the new set to be generated can be prioritized.

(6) If a group management performance value is used as parent selection criterion information,
You can select a parent set based on the excellence of each set. This increases the chances of a good parent set choosing,
As a result, the probability that a better new set is generated can be increased.

(7) If the number of similar sets is used as parent selection criterion information,
The uniqueness of each set can be used as a criterion for selection. This allows
By increasing the probability of selecting a set pair having different characteristics from each other, the diversity of a new set to be generated can be ensured.

(8) It is desirable that the parent selection condition in the parent selection means be modified according to the progress of the search. For example, a plurality of parent selection conditions can be prepared and switched according to the progress, or the reference value of the parent selection conditions can be changed according to the progress. In this way, the probability of generating an excellent new set can be increased.

(9) The parameter selection means selects a parameter based on parameter selection criterion information for increasing the generation probability of an excellent new set. This can increase the probability that an excellent new set will be generated.

(10) If a difference between parameter values is used as parameter selection reference information, similarity between parameters can be used as a selection reference. Thereby, it is possible to give priority to the diversity of the new set to be generated, or to give priority to the convergence of the new set to be generated.

(11) If the degree of association with the elevator use situation is used as the parameter selection criterion information, it is possible to increase the probability that a parameter having a high degree of association with the elevator use state will be selected, and increase the probability of generating a better new set. Can be

(12) If the degree of association with the contents of the performance evaluation value is used as the parameter selection criterion information, the probability that a parameter having a high degree of relation to the contents of the performance evaluation value is selected is increased, and a better new set is generated. The probability of being done.

(13) It is desirable that the parameter selection condition in the parameter selection means be modified according to the progress of the search. For example, a plurality of parameter selection conditions can be prepared and switched according to the progress, or the reference value of the parameter selection conditions can be changed according to the progress. In this way, the probability of generating an excellent new set can be increased.

(14) The selection probability of each generation method is
It is desirable to correct according to the progress of the search. This makes it possible to make use of the features of both the global search by mutation and the local search by crossover, thereby improving search efficiency.

(15) In that case, it is desirable that the selection probability is corrected by the probability correction means, for example, according to the success index. According to this, the progress of the search can be determined from the degree of convergence of the search, and the selection probability suitable for the progress can be set.

(16) Further, in order to achieve the above object, the search device according to the present invention comprises: a storage unit for storing a plurality of sets; and two sets selected as parents from the storage unit. Numerical exchange means for generating two new sets that partially inherit the properties of the parent by exchanging some parameter values with each other; and one of the one set selected as the parent from the storage means. New value replacement means for generating one new set partially inheriting the properties of its parent by replacing the parameter values of the part with new randomly generated values, and said numerical value exchange and said new value replacement Generating method selecting means for selecting stochastically, evaluation means for obtaining an execution result when the group management algorithm is executed using the new set as a group management performance value, Adding means for additionally storing only a good new set satisfying the additional condition in the storage means; deleting means for deleting a bad set satisfying a predetermined deletion condition from the storage means; and storing and improving the excellent set stored in the storage means. Extracting means for extracting the optimum set from the plurality of sets based on the group management performance value; and performing a search based on a change in a precondition of a search given at the start of the search for the optimum set. Re-search determination means for instructing the apparatus to search again for the optimum set.

According to the above configuration, set generation (crossover) by numerical value exchange, set generation (mutation) by new value replacement,
Are randomly selected to generate a new set. Of the generated new sets, only the new sets that satisfy a predetermined additional condition are stored in the storage unit. On the other hand, the bad set is deleted in the storage means.

When such a cycle is sequentially repeated, only excellent sets are accumulated in the storage means, and the optimum set is extracted therefrom. And each numerical value which comprises the optimal set is substituted for each parameter of a group management algorithm.

Therefore, according to the present invention, it is possible to efficiently search for the best set or a set having a content very similar to the best set. That is, the amount of calculation and the number of simulations can be reduced, so that the search can be speeded up.

(17) The search device according to the present invention desirably further includes additional condition correction means.

(18) The additional condition is determined based on, for example, the group management performance value of each set stored in the storage means, and the additional condition is gradually set stricter. In this way, only a good new set can be always stored in the storage means, and unnecessary processing can be reduced, and the probability of generating a good new set can be increased.

(19) The above-mentioned deletion means performs deletion based on, for example, a group management performance value. By doing so, it is possible to raise the excellence of the stored plurality of parent sets as a whole, leaving only the excellence set.

(20) In addition, the above-described deletion unit performs deletion based on, for example, the distance between sets. In this way, with respect to a plurality of sets in the storage means, it is possible to avoid a state in which similar sets are duplicated, and to ensure the diversity of sets.

(21) The search device according to the present invention preferably further includes an initial setting unit. If the initial setting is performed using an initial set group that matches the search conditions as much as possible, the search time can be reduced.

(22) The initial setting means desirably includes the first mode and the second mode.
Mode. In the first mode, a plurality of sets prepared in advance are used as an initial set group, and in the second mode, a plurality of sets improved in the previous search are used as the initial set group. By selecting an appropriate mode according to the state at the start of the search, the convergence of the search can be expedited.

(23) The search device according to the present invention preferably further includes a search end determination unit. This means determines the end of the search when the search progresses to a state where sufficient improvement can be expected. Avoids exiting with insufficient search,
Moreover, useless search can be avoided.

(24) The number of evaluated sets is related to the number of executions of the optimization cycle, and can be used as a termination criterion.

(25) The number of added sets indicates the degree of improvement in the storage unit, and can be used as a termination determination criterion.

(26) The success index is a ratio between the number of evaluated sets and the number of added sets, and indirectly expresses the convergence of the search. Therefore, the success index can be used as an end determination criterion.

(27) Since the set-to-set distance expresses the similarity of the plurality of sets as a whole in the storage unit, it can be used as an end determination criterion.

(28) Preferably, the prerequisites for the search include elevator specification data representing the specifications of a plurality of elevators and traffic flow specification data on passenger traffic. The re-search determination means determines re-search based on a change in the precondition given at the start of the search. According to this, the optimum set can be automatically searched under the new condition. Here, the prerequisites are elevator specifications and traffic flow specifications, and may include performance reference values / control reference values.

(29) In the present invention, the group management performance value can be further stored in the storage means.

(30) In the present invention, a target value setting device can be connected to the search device in order to set a target value for the search. The control target can be freely set, and the optimum set that matches the set target can be searched.

(31) In the present invention, when a new set is evaluated using a simulation-dedicated device, a simulation device is connected to the search device in addition to the group management device. Here, the simulation device includes the same group management algorithm as the group management algorithm included in the group management device. Then, the evaluation unit sets the simulation execution result as a group management performance value. If a simulation device is used, a new set can be evaluated without interrupting group management.

(32) In the present invention, for example, the group management device is arranged in the same building together with the search device (and the simulation device). However, when it is necessary to arrange a search device (and a simulation device) separately from the group management device,
The group management device and the search device are connected by a communication line.
If one search device (and simulation device) is shared by a plurality of group management devices, the system cost can be reduced.

(33) In the present invention, a simulation can be performed using a group management device as an actual device connected to the search device. Since a simulation device is not required, the system cost can be reduced.

(34) In the present invention, when the search device is arranged separately from the group management device, the group management device and the search device are connected by a communication line. If one search device is shared by a plurality of group management devices, the system cost can be reduced.

(35) In the present invention, a simulation device and a search device are connected, and the present invention can be extended and applied to evaluation of a group management algorithm and the like.

-Relationship between the present invention and GA- Genetic algorithms (GA) have been described in various documents (for example, "Measurement and Control" (Vol. 32, No. 1, January 1993)). "Current status and issues of genetic algorithms" described in). The basic GA is initialized,
Includes a series of cycles of parent selection, crossover, mutation, and alternation.

In the 34th Automatic Control Federation Lecture Meeting (November 20-22, 1991), "Solution of elevator group management car assignment problem for definite calls by Genetic Algorithm"
A system in which A is applied to elevator group management is described.

Here, in this conventional system, an elevator is optimally assigned to a "permutation of calls" using a GA. Therefore, although the present invention and the conventional system are common in that they are systems related to GA, the search target is different, and the basic configuration is greatly different.

In short, the present invention provides a new technique for searching for a new optimal parameter value set, not simply using GA. It is revealed by the unique configuration of the present invention due to the unique nature of the parameter value set.

BRIEF DESCRIPTION OF THE DRAWINGS FIG. 1 is a block diagram showing an overall configuration of a system according to a first embodiment of the present invention.

FIG. 2 is a diagram showing a configuration of a search device composed of a microcomputer.

FIG. 3 is a diagram showing a configuration inside the RAM 10C of FIG.

FIG. 4 is a diagram showing a configuration inside the ROM 10B of FIG.

FIG. 5 is a diagram showing a configuration of the elevator specification data (ELS).

FIG. 6 is a diagram showing a configuration of the traffic flow specification data (TRS).

FIG. 7 is a diagram showing a configuration of group management performance data (PRF).

FIG. 8 is a diagram illustrating a configuration of the parameter value set (EPS).

FIG. 9 is a flowchart illustrating the contents of the control program according to the first embodiment.

FIG. 10 is a flowchart illustrating a search command program according to the first embodiment.

FIG. 11 is a flowchart illustrating a search main program according to the first embodiment.

FIG. 12 is a flowchart illustrating a search start determination program according to the first embodiment.

FIG. 13 is a flowchart illustrating an initialization program according to the first embodiment.

FIG. 14 is a flowchart illustrating a new set generation program according to the first embodiment.

FIG. 15 is a flowchart illustrating an evaluation program according to the first embodiment.

FIG. 16 is a flowchart illustrating an additional program according to the first embodiment.

FIG. 17 is a flowchart illustrating a deletion program according to the first embodiment.

FIG. 18 is a flowchart illustrating an additional reference value correction program according to the first embodiment.

FIG. 19 is a flowchart illustrating a search end determination program according to the first embodiment.

FIG. 20 is a flowchart illustrating an optimal set extraction program according to the first embodiment.

FIG. 21 is a block diagram illustrating a system configuration according to the second embodiment.

FIG. 22 is a diagram illustrating the configuration of the RAM according to the second embodiment.

FIG. 23 is a flowchart illustrating an additional reference value correction program according to the second embodiment.

FIG. 24 is a flowchart illustrating a search start determination program according to the second embodiment.

FIG. 25 is a flowchart illustrating an initialization program according to the second embodiment.

FIG. 26 is a flowchart illustrating a deletion program according to the third embodiment.

FIG. 27 is a flowchart illustrating a search end determination program according to the fifth embodiment.

FIG. 28 is a flowchart illustrating a search end determination program according to the sixth embodiment.

FIG. 29 is a flowchart illustrating an optimum value extraction program according to the seventh embodiment.

FIG. 30 is a block diagram illustrating a system configuration according to the eighth embodiment.

FIG. 31 is a flowchart illustrating a search main program according to the eighth embodiment.

FIG. 32 is a block diagram illustrating a system configuration according to the ninth embodiment.

FIG. 33 is a flowchart showing a search main program according to the ninth embodiment.

FIG. 34 is a flowchart illustrating an appearance rate correction program according to the ninth embodiment.

FIG. 35 is a flowchart illustrating an appearance rate correction program according to the tenth embodiment.

FIG. 36 is a block diagram illustrating a system configuration according to the eleventh embodiment.

FIG. 37 is a flowchart illustrating an arithmetic main program according to the eleventh embodiment.

FIG. 38 is a flowchart illustrating a part of the new set generation program according to the eleventh embodiment.

FIG. 39 is a flowchart illustrating a selection condition correction program according to the eleventh embodiment.

FIG. 40 is a flowchart illustrating a selection condition correction program according to the twelfth embodiment.

FIG. 41 is a flowchart illustrating a part of the new set generation program according to the thirteenth embodiment.

FIG. 42 is a flowchart illustrating a selection condition correction program according to the thirteenth embodiment.

FIG. 43 is a flowchart illustrating a selection condition correction program according to the fourteenth embodiment.

FIG. 44 is a flowchart illustrating a part of the new set generation program according to the fifteenth embodiment.

FIG. 45 is a diagram illustrating an appearance rate for each parameter in the fifteenth embodiment.

FIG. 46 is a flowchart illustrating a part of the new set generation program according to the sixteenth embodiment.

FIG. 47 is a diagram illustrating the system configuration of the seventeenth embodiment.

FIG. 48 is a diagram illustrating a system configuration according to the eighteenth embodiment.

FIG. 49 is a diagram illustrating a system configuration of the nineteenth embodiment.

FIG. 50 is a conceptual diagram showing the optimal set search method according to the present invention.

Embodiment [Basic Principle] As described above, the group management algorithm includes a plurality of types of parameters. In order to efficiently manage a plurality of elevators in a group, it is necessary to search for an optimal combination of parameter values according to traffic conditions. An apparatus for this is an optimum set search apparatus, and FIG. 50 shows the basic principle of the search apparatus according to the present invention. Note that, as described above, a combination (numerical sequence) of parameter values is referred to as a “parameter value set” or simply “set”.

In FIG. 50, the optimum set is searched for by repeatedly generating a new set and selecting a good set. Hereinafter, this will be described specifically.

First, initialization is performed on the storage unit A2 (A
1). For example, a plurality of initial sets prepared in advance are stored in the storage unit (A2).

Next, a new set is generated (A4). Here, the new set is generated by randomly selecting any generation method from numerical exchange (crossover) and new value substitution (mutation). When the crossover is selected, two sets (parent set pairs) are extracted from the storage unit A2, and
Some parameter values are exchanged between the two sets to generate two new sets. If a mutation is selected, one set (parent set) is fetched from storage A2, some parameter values in that set are replaced with new randomly generated values, and one new set is Generated.

Note that the selection of the generation method, the selection of the parent set, and the selection of the parameter for replacing the numerical value are basically performed randomly. In addition, you can also set the selection conditions,
In addition, selection probability can be weighted for each selection element.

Next, the generated new set A5 is evaluated (A6). That is, the group management algorithm to which the new set is substituted is executed virtually or actually, and the execution result is obtained. The execution result is obtained as a “group management performance value” indicating the performance of the new set, and the new set having the excellent group management performance value is stored in the storage unit A2 (A
8). The bad set is culled without being stored (A
9) Or deleted after storage (A10). Due to the selection of such excellent sets (A7), only excellent sets are always accumulated in the storage unit A2.

When the above optimization cycle is sequentially repeated, the storage unit
A plurality of sets A3 stored in A2 are gradually selected and further improved.

Then, finally, one of the plurality of stored sets A3 is extracted as the optimal set (A11), and the optimal set is supplied to the group management algorithm, and the group management is performed.

According to the above-described optimal set search method, a child set that inherits the good properties of the parent set can be efficiently generated. That is, it is possible to increase the probability that an excellent child set can be generated from an excellent parent set, and to quickly search.

Here, it is possible to employ only one of the crossover and the mutation, but desirably, both can be selected at random, and both of the convergence and diversity are appropriately determined for a plurality of accumulated sets. It is better to be satisfied.

[First Embodiment] -Description of Configuration- Figs. 1 to 20 are diagrams showing a first embodiment of an elevator group management system according to the present invention. FIG. 1 shows the overall configuration of the system.
A well-known simulation device 2, a search device 10,
including.

The group management device 1 is composed of a microcomputer. In this example, the group management device 1 is installed in a ten-story office building.
Group elevators are managed. As described above, the group management device 1 has a group management algorithm including a plurality of types of control parameters (see FIG. 9).

Four car control devices 1A to 1D are connected to the group management device 1 via a communication cable. Car control device 1A ~
Each 1D is composed of a microcomputer,
Various controls are performed for the elevator in charge.
Each of the car control devices 1A to 1D has a car call registration function, an operation control function, a door control function, a display control function, and the like.

Here, the car call registration function is a function of registering a car call when it occurs in the storage unit. The operation control function is a function of controlling traveling of the elevator, stopping, determining an operation direction, and the like in order to cause the elevator to respond to a call to be answered (a car call, an assigned hall call).
The door control function is a function of controlling the opening / closing timing and the door opening time of the elevator door and the landing side door. The display control function is a function of turning on a hall lantern to notify a waiting passenger at the landing of an assigned elevator in advance or notifying the arrival of the elevator by blinking of the hall lantern.

Each of the car control devices 1A to 1D transmits a signal indicating an operation state (a car position, an operation direction, a door opening / closing state, a car call, and the like) to the group management device 1. On the other hand, the group management device 1 transmits signals representing various commands (a command for allocating a hall call, a reference value DB for full passage, a set value of a door opening time, and the like) to the car control devices 1A to 1D.

The group management device 1 outputs to the search device 10 a search condition signal 1a indicating a condition for searching for the optimal set. The search condition signal 1a includes "elevator specification data" required for simulating the group management elevator system on a computer and "traffic specification data" required for simulating a traffic flow in a building on a computer. And "search command data" for commanding a search for an optimal set.

The above-mentioned elevator specification data includes, for example, data indicating the number of elevators, speed, capacity, stop floor, door type, and presence / absence of additional operations such as power saving operation and operation at work. For example, when the traffic flow in the building is indirectly specified, the above-mentioned traffic flow specification data includes data combining characteristic values such as the total number of passengers per hour and the traffic ratio between floors, and floor-by-floor direction. And data combining characteristic values such as the number of passengers per unit time. On the other hand, if the traffic flow in the building is directly specified, the passenger data (occurrence time, occurrence floor, Destination floor, etc.).

The simulation device 2 is configured by a microcomputer, and has the same group management algorithm as the group management device 1. The simulation device 2 receives a simulation condition signal 13a including elevator specification data, traffic flow specification data, and a parameter value set. Based on the signal 13a, the simulation apparatus 2 virtually operates a plurality of elevators under the same conditions as the actual conditions under the group management algorithm. And
After execution, statistical results (average waiting time,
Group management performance data indicating the long wait rate, etc.) is output as the group management performance value signal 2a.

The search device 10 is configured by a microcomputer, and searches for an optimal set as described above.

In such a search device 10, the storage unit 11 stores a plurality of parameter value sets, and also stores the group management performance data in association with each set. In addition,
The output signal 11a from the storage unit 11 is composed of parameter value set data and group management performance data.

The generation unit 12 generates a new set by the above “crossover” and “mutation”. The new set is
This generation unit 12 is evaluated until evaluated by an evaluation unit 13 described later.
Is stored temporarily. Note that the generating unit 12 outputs a new set signal 12a.

The evaluation unit 13 includes a search condition signal 1a and a new set signal 12a.
And generates a simulation condition signal 13a based on the above, and outputs it to the simulation device 2. Further, after the execution of the group management simulation, the evaluation unit 13
Group management performance signal output from the simulation device 2
Based on 2a, an evaluation result signal 13b is created and output to the adding unit 15.

The additional reference value memory 14 stores an additional reference value for determining whether the new set evaluated as described above is to be additionally registered in the storage unit 11 or to be eliminated. The additional reference value memory 14 outputs an additional reference value signal 14a.

The adding unit 15 creates a performance evaluation value for additional registration determination from the group management performance data included in the evaluation result signal 13b, and compares this with an additional reference value. If the performance evaluation value is better than the additional reference value, a signal 15a including the new set and its group management performance value is created and output to the storage unit 11. Thereby, an excellent new set is additionally registered in the storage unit 11.

The deletion unit 16 obtains a performance evaluation value for deletion determination for each set based on the group management performance data when a predetermined condition regarding the set registration status is satisfied. Then, the deletion unit 16 selects a set having a poor performance evaluation value, and outputs a deletion command signal 16a indicating the set number. Thus, the registration of the specified set is peripherally performed from the storage unit 11.

The end determination unit 17 determines whether to end the search,
When the search end is determined, the search end signal 17a is generated by the generation unit 1.
Output to 2. Thus, the generation of the new set ends.

The additional reference value correction unit 18 corrects the additional reference value stored in the additional reference value memory 14 using the correction signal 18a. The correction degree is determined based on the group management performance data of each set in the storage unit 11.

The re-search determination unit 19 monitors the search command signal 1a, and outputs a re-search command signal 19a for re-executing the search for the optimal set when the elevator specification or the traffic flow specification changes. When this signal 19a is output, the search end command by the search end signal 17a is invalidated and the search is started again from the beginning, and the search is started again from the beginning even during the search.

The extraction unit 20 obtains a performance evaluation value for optimal set determination based on the group management performance data of each set in the storage unit 11,
The best set is extracted. That is, an optimal set is extracted. The signal 20a output from the extraction unit 20 includes an optimal set, elevator specification data,
It is composed of traffic flow specification data and search state data.

The initial setting unit 21 includes a plurality of initial set groups, and is used in the initial setting among a plurality of pre-stored initial set groups according to the search condition signal 1a or the re-search instruction signal 19a at the start of the search. An appropriate set group is specified and output to the storage unit 11.

FIG. 2 shows a hardware configuration of the search device 10 shown in FIG. In FIG. 2, the search device includes a microprocessor 10A, a read-only memory (ROM) 10
B, a readable / writable memory (RAM) 10C, an input interface circuit 10D, and an output interface circuit 10E. Here, the ROM 10B stores the microprocessor 10
It stores a search program describing the operation procedure of A and fixed data. Further, the RAM 10C stores the operation result (operation data) of the microprocessor 10A, the contents (input data) of the search condition signal 1a and the group management performance value signal 2a input from the outside, the simulation condition signal 13a output to the outside, Contents of set signal 20a (output data)
And is stored.

FIG. 3 shows the storage contents of the RAM 10C of FIG.
FIG. 4 shows a fixed data portion of the storage content of the ROM 10B.

In FIG. 3, ELS is data indicating an elevator specification, TRS is data indicating a traffic flow specification, and SCM is data indicating a search command. These input data are
It is included in the search condition signal 1a shown in FIG.

FIG. 5 shows a specific configuration of the elevator specification data ELS. In the example shown in FIG. 5, the number of elevators is four.
Stand, speed 120m / min, capacity 20 people, stop floor 1 at the lowest floor
On the floor, the top floor has 10 stops on the 10th floor, and the door width is set to 1000 mm. In the priority assignment operation, [2] the ride time priority assignment function, [3] the power saving assignment function, [4] the proximity elevator priority assignment function, and [5] the light load elevator priority assignment function are all “ "Valid" and [6] the specific elevator priority assignment function is set to "invalid". Among the various driving patterns, [11] and [14]
[2], [15] Up-peak operation, [13], [16] Down-peak operation, and [17] Distributed standby operation are all set to "valid" and other additional operation [1]
8] The power saving operation is also set to "valid" (although not shown, [8] assignment change operation for a long waiting call, [9] automatic full passage function, and [10] full passage The assignment change operation for a given hall call is basically
Always "valid").

FIG. 6 shows a specific configuration of the traffic flow specification data TRS. In the example shown in FIG. 6, the business hour (14:00
~ 15: 00). For example, based on the result of actually measuring the traffic flow using the group management device 1,
The total number of passengers per hour is 500, the ratio of the traffic volume between the first floor and the ground floor (2nd to 10th floor) to the total traffic volume (= entrance floor traffic ratio) is 80%, and the total traffic volume , The ratio of the traffic volume in the up direction (= up traffic ratio) is set to 50%, and the ratio of the traffic volume in the down direction (= down traffic ratio) is set to 50%.

PRF at the upper left of FIG. 3 is data representing group management performance indicating the excellence of each set, and is a group management performance signal 2a in FIG.
Is equivalent to

FIG. 7 shows a specific configuration of the group management performance data PRF. In this example, the group management performance data PRF includes an average waiting time AWT, a long waiting time RLW, a maximum waiting time MWT, and a forecast miss rate R.
PE, forecast change rate RPC, packed passing rate RBP, average riding time
ABT, Maximum Ride Time MBT, Power Consumption PWC, Proximity Elevator Response Rate RNR (Ratio of elevators located near the button when operating a landing button), Light Load Elevator Response Rate RLR When registered, a light load elevator is allocated) and a specific elevator response rate RSR (when a hall call is registered, a specific elevator is allocated).

Returning to FIG. 3, P at the upper right of FIG. 3 is data representing the number of sets (also referred to as excellent sets) registered in the storage unit 11, and EPS (1) to EPS (Pmax) are set numbers 1
Data representing sets from to (Pmax), PRE (1) to PRE
(Pmax) is data representing a group management performance value corresponding to EPS (1) to EPS (Pmax). Here, the set number P, the set data EPS (1) to EPS (Pmax), and the group management performance data
PRE (1) to PRE (Pmax) correspond to the signal 11a in FIG. Here, Pmax described later is data representing the maximum value of the number of sets that can be registered.

FIG. 8 shows a specific configuration of the parameter value set as an example. In FIG. 8, this set consists of 25
It consists of different types of control parameters. That is, each set data EPS (1) to EPS (Pmax) shown in FIG.
It is configured as shown in FIG. Note that the group management performance data PRE (1) to PRE (Pmax) in FIG. 3 have the same configuration as the group management performance data PREF (see FIG. 7 for the specific configuration) in FIG.

In the middle part on the right side of FIG. 3, Pn is data representing the number of new sets generated, NPS (1) to NPS (Nmax) are set numbers 1 to
Data representing a new set of Nmax. Number of new sets P
n and the new sets NPS (1) to NPS (Nmax) are shown in FIG.
Signal 12a. Here, Nmax described below is data representing the maximum value of the number of new sets that can be generated.

SIM at the upper right of FIG. 3 is output data corresponding to the simulation condition signal 13a of FIG. 1, and is composed of evaluation set data NPSX, elevator specification data ELSX, and traffic flow specification data TRSX. The evaluation set data NPSX is
This is data indicating the contents of a new set whose group management performance is to be evaluated by simulation, and has the same configuration as the EPS in FIG. The elevator specification data ELSX and the traffic flow specification data TRSX are data representing the elevator specification and the traffic flow specification when the simulation is performed, and have the same configurations as the ELS in FIG. 5 and the TRS in FIG. 6, respectively.

RES at the upper left of FIG. 3 is data corresponding to the evaluation result signal 13b of FIG. 1, and includes the number of evaluations NE, the evaluation set NPSY, and
It consists of group management performance data PRFY. The number of evaluations NE is data representing the cumulative number of evaluations. Also, the evaluation set NP
SY is data representing a new set after evaluating group management performance by simulation, and has the same configuration as EPS in FIG. Further, the group management performance data PRFY is data representing a group management performance value obtained by simulation, and FIG.
It is configured similarly to the PRF.

BX is data representing an additional reference value for determining whether or not to add and register an evaluated new set, and is data corresponding to the additional reference value signal 14a in FIG.

RAP is data corresponding to the additional registration signal 15a in FIG.
It consists of an additional registration number NR, an evaluation set NPSZ, and group management performance data PRFZ. The number of additional registrations NR is data representing the number of times of determining additional registration. The evaluation set NP
SZ is data representing an excellent new set to be registered in the storage unit 11, and has the same configuration as the EPS in FIG. The group management performance data PRFZ is data representing group management performance when performing group management simulation using the evaluation set NPSZ, and has the same configuration as the PRF in FIG.

The RP in the middle part on the left side of Fig. 3 is a set of registered P EPS
Regarding (1) to EPS (P), this is data indicating the number of the set whose registration is to be deleted as an inferior set, and is data corresponding to the delete command signal 16a in FIG.

FLAG is data (search permission flag) indicating whether to continue the optimum set search or to end the search, and is data corresponding to the search end signal 17a in FIG.

CBX is data for newly rewriting the additional reference value BX, and is data corresponding to the correction signal 18a in FIG.

STR is data instructing re-execution of the optimal set search, and is data corresponding to the re-search instruction signal 19a in FIG.

The BPD is output data corresponding to the optimal set signal 20a in FIG. 1, and is composed of an optimal set BPS, elevator specification data ELSY, traffic flow specification data TRSY, and search state data SS. The optimal set BPS is a registered set,
This is the set having the best performance evaluation value, and has the same configuration as the EPS of FIG. In addition, elevator specification data ELSY,
The traffic flow specification data TRSY is data representing elevator use and traffic flow use when a group management simulation was performed using the optimal set BPS.
It has the same configuration as S and the TRS in FIG. Further, the search state data SS is data representing a search state when the optimum set BPS is selected, and in this embodiment, a value indicating the number of evaluations NE is set.

GPS0 at the lower left of FIG. 3 is data corresponding to the initial setting signal 21a of FIG.
S (1) to IPS (Pk) and multiple group management performance data PR
It is composed of I (1) to PRI (Pk). The initial set number Pk is
This is data representing the number of sets at the start of the search, and is usually set to the same value as the determination value Pe for determining the end of deletion. Also, a plurality of initial sets IPS (1) to IPS (Pk) are prepared in advance as a set group at the start of the search, and are shown in FIG.
It has the same configuration as EPS. Group management performance data PRI (1)-
PRI (Pk) is data representing group management performance when the group management simulation apparatus is implemented using the initial sets IPS (1) to IPS (Pk), and has the same configuration as the PRF in FIG.

VPD (1) to VPD (Pmax) on the right side of FIG. 3 are performance evaluation values for deletion determination, and VPE (1) to VPE (Pmax) are additional reference values used when correcting the additional reference value BX. The performance evaluation value for setting, VPS (1) to VPS (Pmax), is the optimal set BPS
3 is a performance evaluation value for determining an optimal set, and VPN in the lower left portion of FIG. 3 is a performance evaluation value for additional determination used when determining whether to additionally register an evaluation set NPSY. In this embodiment, the average waiting time AWT extracted from the group management performance data is directly substituted for each performance value.

NP is data representing a set number for which group management performance is to be evaluated among the new sets NPS (1) to NPS (Nmax).

WVPE is the worst value of performance evaluation values, BVPE
Is the data representing the best value among the performance evaluation values, RC is the worst value
A search counter for counting the set number used when searching for WVPE and the best value BVPE, and BP is data representing the number of a registered set having the best value BVPE.

PS1 is a first parent set number indicating a parent set number for generating a new set, PS2 is a similar second parent set number, PX is a parameter number (position) to be subjected to "crossover" or "mutation" ), CR is data representing the crossover selection probability (appearance rate), and MR is data representing the mutation selection probability (appearance rate).

In FIG. 4, Pmax is data representing the maximum value of the number of sets that can be registered, and Nmax is data representing the maximum value of the number of new sets that can be generated. In the first embodiment, Pmax = 50.
And Nmax = 20.

NEa is a search end determination value used when determining whether the search for the optimal set has converged by the search frequency NE. In the first embodiment, NEa is set to 1,000 times.

Also, Ps is a deletion start determination value used when determining whether or not to delete the inferior set based on the set registration number P, and Pe is a deletion end determination used when determining whether to end the inferior set deletion process. In this embodiment, the values are set to Ps = 50 and Pe = 30, respectively.

AVPE is data indicating a correction value to be added to the worst value WVPE of the performance evaluation value when setting the additional reference value CBX. That is, the additional reference value is corrected as a value obtained by adding the correction value AVPE to the worst value WVPE. As the correction value, generally, a value of 0 second or more is set. In the first embodiment, AVPE = 1 second. GPS1 to GPS4 are an initial set group corresponding to normal driving (business hour), driving at work, up-peak driving, and down-peak driving. Each initial set group GPS1 ~ GPS4
Has the same configuration as the initial set group GPS0 in FIG.

-Description of Operation- Next, the operation of the first embodiment will be described with reference to FIGS.

FIG. 9 shows a main configuration of a control program of the group management device 1. This control program includes a group management algorithm, and the group management device 1 performs control based on the control program. Note that the group management algorithm itself is known.

In FIG. 9, at step 221, a hall call registration program is executed. Specifically, a hall call generated when the hall button is operated by the passenger is registered in the memory.
If any elevator responds to the call, the registration of the call is deleted.

In step 222, the assignment program is executed. Specifically, the assignment evaluation function of the above equation [1] is used,
An assignment evaluation value is calculated for each elevator. Then, the elevator having the minimum evaluation value is assigned to the call. In this step, in addition to the basic assignment calculation using the above evaluation function, [2] ride time priority assignment function, [3] power saving assignment function, [4] proximity elevator priority assignment function, [5] light weight assignment function Processing based on the load elevator priority assignment function and [6] the specific elevator priority assignment function are also included.

In step 223, an assignment change program is executed. Specifically, it is detected that the service of the hall call assigned as described above has deteriorated, and an assignment for rescue is performed. In this step,
[8] A process based on an assignment change operation for a long waiting call and [10] an assignment change operation for a hall call that has passed through a packed room is also included.

In step 224, the in-service driving program is executed. Specifically, [11] the operation mode is selected / eliminated according to the selection / elimination conditions for driving at work, and
When the on-the-job driving is selected, the operation control according to [14] on-the-job driving operation is performed.

In step 225, an up-peak operation program is executed. Specifically, [12] the operation mode is selected / eliminated in accordance with the up-peak operation selection / elimination condition, and when the up-peak operation is selected,
[15] Operation control according to the up-peak operation is performed.

In step 226, a down peak operation program is executed. Specifically, the operation mode is selected / eliminated according to the above-mentioned [13] Down-peak operation selection / elimination conditions, and when down-peak operation is selected, [16] down-peak operation is performed. Operation control is performed.

In step 227, a distributed standby operation program is executed. More specifically, the distributed standby operation is selected when none of the start-up operation, the up-peak operation, and the down-peak operation is selected. When the distributed standby operation is selected, the operation control according to [17] Distributed standby operation is performed.

In step 228, a power saving operation program is executed. Specifically, in order to achieve power saving while considering the operation service status, the operation control is performed while increasing or decreasing the number of operating units according to [18] power saving operation.

In the last step 229, the output program is executed. Specifically, the [9] packed passage reference value DB required for the [9] automatic packed passage function is sent to the four car control devices 1A to 1D connected to the group management device 1. Each of the car control devices 1A to 1D determines whether or not the car is full, based on the car load and the packed passage reference value DB, and if the car is full, automatically passes the call generation floor. Here, the packed passage reference value DB has a large effect on the group management performance, and is therefore treated as a control parameter to be searched.

Note that the entire group management program (including the control program in FIG. 9 and the search command program in FIG. 10 described below) is repeatedly executed periodically (for example, every 100 milliseconds).

FIG. 10 shows a search command program that the group management device 1 has. This program is a program for issuing a search instruction to the search device 10.

In FIG. 10, step 231 is executed when an optimal set has already been determined for any traffic flow. The optimal set signal 20a is fetched from the search device 10 and stored in the memory of the group management device 1. The optimal set BPS is stored in association with the traffic flow TRS. In that case, the search state data SS included in the optimal set signal 20a is also stored.

In steps 232 and 233, the progress of the search is determined based on the search state data SS. If the search state data SS
If it is equal to NEa (= 1,000), the search has been completed, so in step 234, the traffic flow for searching the optimal set is specified, and the specification data TRS of the traffic flow and the elevator specification data ELS are used. , A new search condition signal 1a composed of the search command data SCM set to “1” and output.

On the other hand, if it is determined in step 233 that the search state data SS is 1 or more, it indicates that the search has already been started, and in step 235, the search command data SCM in the search condition signal 1a The value is rewritten to "0" and a new search condition signal 1a is output.

In order to search for an optimal set for each traffic flow, four types of traffic flows corresponding to normal driving (business hour), driving at work, up-peak driving, and down-peak driving are sequentially selected.

In step 234, the traffic flow specification data TRS is created based on the actual measurement result of the group management device 1.
A well-known traffic measurement device is connected to the group management device 1, the collected traffic condition data (number of passengers, number of calls, etc.) is totaled, the data is input to the group management device 1, and the traffic flow is The specification data TRS may be created.

Further, in the search device 10, when the mode in which the optimum set is output at any time during the search is adopted, how much the optimum set obtained up to now is determined according to the search state data SS stored in step 231. You can judge whether it is reliable. For example, search state data
If the SS indicates the initial stage of the search, a set that has been used in the past can be used as the actually used set without using the set input from the search device 10. This can prevent the group management performance of the system from being lowered. In addition, if the search state data SS indicates the middle stage or the final stage of the search, the optimal set output from the search device 10 can be determined to be highly reliable. Before the end, the group management performance can be improved.

FIG. 11 shows a search program (main program) stored in search device 10. This program is stored in the ROM 10B.

11, in step 25, a re-search determination program having the function of the re-search section 19 in FIG. 1 is executed. In step 26, the search start determination program is executed to determine whether it is time to restart the search for the optimal set. The re-search determination method will be described with reference to FIG.

In FIG. 12, in step 261, the search device 10 inputs the search condition signal 1a from the group management device 1, and stores the elevator specification data ELS, the traffic flow specification data TRS, and the search command data SCM in the RAM 10C. And the next step 262
Then, it is detected that the search command data SCM has changed from "0" to "1", and if it is detected, the search start flag STR is set to "1" in step 265. On the other hand, when it is determined that the search command data SCM has not changed from “0” to “1”, in a detection step 263, it is determined whether the elevator specification data ELS is different from the elevator specification data ELSX searched so far. And step 264
It is determined whether or not the traffic flow specification data TRS is different from the traffic flow specification data TRSX searched so far. ELS and ELSX
Is different, or if TRS is different from TRSX, in step 265, the search start flag STR is set to “1”. Otherwise, in step 266, the search start flag STR is set to “0”.

It is to be noted that the determination of the re-search in Steps 263 to 265 is that if the prerequisites for the search have changed compared to before, and there is a high possibility that the currently registered optimal set is no longer optimal, the optimal set is re-executed. It is necessary to search for. For example, when the traffic flow in the building changes due to the change of tenants, or when a part of the group management algorithm is changed to improve the function, the re-search is executed.

Returning to FIG. 11, in step 27, it is determined based on the processing result of step 26 whether or not it is necessary to perform the initial setting again. Here, STR = "0" means that the search is in progress. On the other hand, STR ≠ “0” means that the search is terminated halfway and the search is restarted from the beginning, or that the search is started again after the search is completed. Therefore, STR =
When it is “0”, the process proceeds to the generation program 29 as it is,
If STR ≠ “0”, the initialization program in step 28 is executed, and after initializing various data, the process proceeds to the generation program 29.

Here, the operation of the initial setting program 28 will be described with reference to FIG.

In FIG. 13, in step 281, an initial data group suitable for a designated traffic flow is selected from a plurality of types of initial data groups stored in advance. Here, each initial data group is composed of an initial set number Pk, Pk initial sets, and Pk group management performance data.

For example, when the normal time zone is designated by the traffic flow specification data TRS, a plurality of initial data groups GPS1 to GP
An initial data group suitable for normal operation is selected from S4, and the initial data group is registered as an initial setting data group GPSO in FIG. Here, the initial setting data group GPSO in FIG. 3 includes the initial set number Pk and a plurality of initial set IPs.
S (1) to IPS (Pk) and group management performance data PRI (1) to P
RI (Pk)).

The initial set number Pk is set to the same value (= 30) as the deletion end determination value Pe.

Next, in step 282, the initial set number Pk is set to the set number P, and the initial sets IPS (1) to IPS (Pk) are set to the registered set EP.
Group management performance data PRI (1) to P (P) are added to S (1) to EPS (P).
RI (Pk) is added to group management performance data PRE (1) -PRE (P)
Substituted respectively. That is, as shown in FIG.
The initial setting A1 of the storage unit A2 is executed.

Then, in step 283, the number of evaluations NE is set to 0, the number of additional registrations NR is set to 0, the evaluation target set number NP is set to 0, the search permission flag FLAG is set to “1”, the crossover probability CR is set to 1.0, and the mutation probability is set to 1.0. MR is initialized to 0.01 respectively, and this program ends.

Returning to FIG. 11, in step 29, a generation program corresponding to the generation unit 12 in FIG. 1 is executed. First, step 30
It is determined whether to continue the search. If the search permission flag FLAG is "0", the process returns to the search start determination program in step 26, while the search permission flag FLAG
Is “1”, the process proceeds to a new set generation program in step 31.

Hereinafter, the new set generation program will be described with reference to FIG.

In FIG. 14, first, in step 311, it is determined whether or not a new set that has not been evaluated remains. If the evaluation target set number NP is less than the maximum value Nmax, there is a remaining new set that has not been evaluated.
Exit this program 31 immediately to evaluate the new set. On the other hand, the evaluation target set number NP is the maximum value Nm
If it is equal to or more than ax, that is, if all new sets have been evaluated, the process proceeds to step 312, where the number of generated sets Pn is initialized to zero.

In the next step 313, P registration sets EPS (1) to
Group management performance data for EPS (P) PRE (1)-PRE
Average waiting time AWT (1)-AWT from (P)
(P) are taken out and these are evaluated as performance evaluation values for judging the optimum value.
Substitute VPS (1) to VPS (P). The probability (appearance rate) that each set is selected as a parent set is set based on the reciprocals of these performance evaluation values VPS (1) to VPS (P).

Then, the processing of steps 314 to 324 described below is repeated to generate a new set up to the maximum value Nmax.

First, in step 314, the number of generated sets Pn is increased by one. In the next step 315, from 0 to [CR + M
R] (the sum of the crossover rate and the mutation rate) is generated, and if the value of the random number is less than CR (= 1.0), it is "crossover", CR (= 1.0) or more , A generation method is randomly selected, such as “mutation”.
However, the ratio between the crossover rate (crossover selection probability) CR and the mutation rate (mutation selection probability) MR can be modified based on a predetermined standard.

If “crossover” is determined in step 316, the process proceeds to step 317. Here, first, a reciprocal value of the performance evaluation value VPS is assigned to each set as a weighting value. The magnitude of the weight value indicates the magnitude of the probability of being selected. Then, two random numbers are generated within a range in which [0] is the lower limit and [the sum of the reciprocals of the performance evaluation values VPS (1) to VPS (P)] is the upper limit. Then, two sets are selected according to the value of the generated random numbers. Here, it is assumed that the two parent sets (cross set pairs) are a PS1st set EPS (PS1) and a PS2nd set EPS (PS2).

In the next step 318, one random number having a value between 0 and 25 is generated, and a parameter number PX specified by the random number value is selected. The parameter number is 2
This number is a common number among the parent sets, and specifies the parameter position at which the numerical value exchange is performed.

Then, in step 319, the PX-th numerical value in the parent set EPS (PS1) and the PX-th numerical value in the parent set EPS (PU2) are exchanged with each other. As a result, two new sets are generated. Then, the generated two sets are set as a [Pn] -th new set NPS (Pn) and a [Pn + 1] -th new set NPS (Pn + 1).

Finally, in step 320, the value of the number of generated new sets Pn is increased by one.

On the other hand, when it is determined as “mutation” in step 316, the process proceeds to step 321. Here, first, for each set,
The value of the reciprocal of the performance evaluation value VPS is given as a weight value. The magnitude of the weight value indicates the magnitude of the probability of being selected. Then, one random number is generated within a range where [0] is the lower limit and [the sum of the reciprocals of the performance evaluation values VPS (1) to VPS (P)] is the upper limit. Then, one set is selected according to the value of the generated random number.

Further, in step 322, a random number is generated and a parameter number PX to be mutated is selected, as in step 318.

Then, in step 323, between the [minimum value] and the [maximum value] that the parameter specified by the number PX can take,
Generate one random number. PS 1st excellent set EPS (PS
Replace the PXth number in 1) with a random value. The generated set is converted to the Pnth new set NPS (Pn)
Set as

In the next step 324, it is determined whether the required number of new sets have been generated. Here, considering that two new sets are generated at one time due to “crossover”, the generation of the new set ends when Pn + 2> Nmax.
Until then, repeat steps 314 to 324 to
Generate Nmax new sets. When the generation of the new set is completed, in step 325, the set number of the new set to be evaluated first, that is, the evaluation target set number NP is initialized to 1.

“Crossover” is a search method having convergence of the solution, while “mutation” is a search method having a variety of solutions. That is, according to the crossover alone, the direction of the search becomes local, and the possibility of losing the optimal solution increases. However, it is possible to escape from the local search by appropriately using the mutation. In that sense, they are complementary. However, on the contrary, there is a risk that the optimal solution searched for may be destroyed due to the mutation. In that sense,
The two are in competition.

Therefore, in order to minimize the risk of a competitive relationship while using the complementary relationship, in the first embodiment, the mutation rate MR is set to a very small value compared to the crossover rate CR.

In the present embodiment, after generating 20 (= Nmax) new sets, the group management performance of each new set is evaluated. However, other techniques can be applied.

For example, by setting the maximum value Nmax to “1”, each time “crossover” or “mutation” is performed once, the evaluation of the group management performance for the new set can be performed, and the evaluation can be repeated.

Incidentally, when the value of the maximum value Nmax increases, a new set can be generated at a time, so that the calculation time can be reduced. However, in that case, it is necessary to secure a large capacity of the RAM 10C. It is desirable to appropriately determine the value of the maximum value Nmax in consideration of the calculation time and the amount of memory.

Returning to FIG. 11, the evaluation program in step 33 will be described. This evaluation program is based on the evaluation
13 corresponds to 13 in which a new set is given to the simulation apparatus to execute the group management algorithm, and the execution result is obtained. About this evaluation program,
This will be described using 15.

In FIG. 15, first, in step 331, simulation condition data SIM including an evaluation set NPSX, elevator specification data ELSX, and traffic flow specification data TRSX is created. That is, a new set NPS (NP) is set in the evaluation set NPSX, and the elevator specification data ELSX and the traffic flow specification data TRSX include the elevator specification data ELS and the traffic flow specification data T included in the elevator specification signal 1a.
Set RS. Then, in the next step 332, the simulation condition data SIM is converted to the simulation condition signal 13a.
To the simulation device 2 to cause the simulation device 2 to perform a virtual group management operation. Then, in step 333, the process waits for the end of the simulation.

The simulation device 2 performs a simulation according to the simulation condition signal 13a, and outputs the group management performance value signal 2a to the search device 10 when the simulation ends. In step 333, when the group management performance value signal 2a is received, it is determined that the simulation has been completed. move on.

In step 335, the value of the number of evaluations NE is increased by 1 and updated, and the number of evaluations NE and the evaluation set NPSY (= NPS
X) and the evaluation result data RES including the group management performance data PRFY (= PRF) is created. Then, in step 336, the value of the evaluation target set number NP for the new set is increased by 1 and updated.

Returning to FIG. 11, the additional program of step 34 is shown in FIG.
And determines whether or not to register a new set (unit number NP). This additional program will be described with reference to FIG.

In FIG. 16, first, in step 341, the average waiting time AWT is extracted from the group management performance data PRFY, and is set as a performance evaluation value VPN for additional registration determination. Then, in step 342, the performance evaluation value VPN is compared with the additional reference value BX, and it is determined whether or not to be registered in the storage unit. If VPN
If ≧ BX, registration is not allowed and this program
To end.

On the other hand, if VPN <BX, the process proceeds to step 343, where the value of the number of additional registrations NR is increased by 1 and updated, and the number of additional registrations NR, the evaluation set NPSZ (= NPSY), and the group management performance Data RAP for additional registration consisting of data PRFZ (= PRFY)
Create Then, in step 344, additional registration is performed as the [P + 1] -th set, and at the same time, the value of the registered set number P is increased by one and updated.

Returning to FIG. 11, the deletion program of step 35
And deletes sets with poor performance evaluation values. The deletion program will be described with reference to FIG.

In FIG. 17, first, in step 351, the number of registered sets P is compared with a deletion start determination value Ps, and it is determined whether it is time to delete a set. If P <
If it is Ps, it is determined that it is not time to delete the excellent set, and the program exits this program 35 immediately. Also, P
If ≧ Ps, it is determined that it is time to delete an excellent set, and the following steps 352 to 359 are repeated to reduce inferior sets until the set number P reaches the deletion end determination value Pe.

In step 352, group management performance data PRE (1) to PRE
(P) Average waiting time AWT (1)-AWT
(P) is taken out, and these are deleted. Performance evaluation value VPD
(1) to VPD (P). And
In step 353, initialization is performed to detect a bad set to be deleted. That is, the search counter RC is set to 1
, The worst value WVPE of the performance evaluation value is set to 0, and the deletion set number RP
Is set to 0.

Then, by repeating the processing of steps 354 to 357, the set having the worst performance evaluation value (set number RP)
Is specified. That is, in step 354, every time a set whose performance evaluation value VPD (RC) is worse than the worst value WVPE up to that point is detected, in step 355, the performance evaluation value VPD
(RC) is set to the worst value WVPE. Also, delete set number
Set the value of the search counter RC in RP. Step 356
In step 357, the search counter RC is advanced by one.
It is determined whether the search for all sets has been completed.

In step 358, the registration of the set having the worst value WVPE (deletion set number RP) is deleted, and the registration of the group management performance data PRE (RP) is also deleted. Further, the update is performed with the value of the registered set number P also reduced by one. Then, for the remaining sets, the set numbers are sequentially added again from 1 and stored again, and the processing of step 358 is terminated.

In step 359, it is determined whether or not the number P of sets after deletion has become equal to or less than a deletion end determination value Pe. here,
If not, the processes in steps 352 to 358 described above are repeated. Then, when P ≦ Pe, the execution of the deletion program ends.

In step 351, in the present embodiment, the number of deletion start determination values Ps is set to 50 and the number of deletion end determination values Pe is set to 30. However, the present invention is not limited to this.

The deletion start determination value Ps may be set within a range that does not exceed the maximum number Pmax of sets that can be stored in the RAM 10C. In addition, if the deletion start determination value Ps is set so that Ps = Pe + 1, a technique of "every additional set is registered, one alternative set is deleted" can be realized. This technique is
This is convenient when the storage capacity of the RAM 10C is not enough.

The deletion end determination value Pe in step 359 means the number of remaining parent sets. If the deletion end determination value Pe is small, it becomes difficult to maintain the diversity of the new set to be generated, so that the probability of generating an excellent new set decreases.
Conversely, if the deletion end determination value Pe is large, the diversity of the new set to be generated can be secured, and as a result, the probability of generating a better new set can be increased. However, since the amount of calculation for generation increases, it is not desirable to make the deletion end determination value Pe too large from the viewpoint of efficient search.

Therefore, the number of combinations of two sets that perform crossover,
It is desirable to set the deletion end determination value Pe by trial and error in some cases according to the type and number of control parameters. In the first embodiment, since Pe = 30, 30 × 29 ÷
2 = 435 combinations are secured.

Returning to FIG. 11, the additional reference value correction program in step 36 corresponds to the additional reference value correction unit 18 in FIG. 1, and corrects the additional reference value BX according to the set registration status of the storage unit 11. The additional reference value correction program will be described with reference to FIG.

In FIG. 18, first, in step 361, the average waiting times AWT (1) to AWT (P) are extracted from the group management performance data PRE (1) to PRE (P), respectively, and are extracted as performance values for setting a reference value. Substitute for the evaluation values VPE (1) to VPE (P). Next, in step 362, the performance evaluation value VPE for setting the reference value is set.
(1) An operation for characterizing the worst value MVPE in VPE (P) is performed. This calculation is performed in step 35 of FIG.
This is the same as the processing in steps 3 to 357. Then, in step 363,
[Worst value WVPE−Correction value AVPE] is calculated to obtain a correction value CBX. In step 364, this correction value CBX is added to an additional reference value.
Modify by assigning to BX.

In this embodiment, the correction value AVPE is fixed at 1 second from the beginning to the end of the search. That is, the additional reference value BX indicating the average waiting time is set so as to decrease by one second. However, other values can be employed.

If the value of the correction value AVPE is increased, the conditions for additional registration gradually become stricter.Therefore, if it is important to obtain as many excellent sets as possible within a limited number of evaluations, use a very large value. Should not be. Conversely, if the correction value AVPE is set to 0 seconds, the possibility of additional registration of many sets having similar characteristics with almost no difference in performance increases. Therefore, it is necessary to appropriately determine the value according to the search condition.

Returning to FIG. 11, the search termination determination program of step 37 corresponds to the search termination determination unit 17 of FIG. 1, and determines whether the search for the optimal set has been completed. This will be described with reference to FIG.

In FIG. 19, in step 371, it is determined whether or not to end the search based on the number of evaluations NE and the search end determination value NEa. If NE <NEa, it is determined that the search has not been sufficiently performed, and in step 372, the search permission flag FLAG is set to “1” to continue the search.
When it is determined that NE ≧ NEa and the search has been sufficiently performed, in step 373, the search permission flag FLAG is set to “0” to end the search.

In this embodiment, the value of the search end determination value NEa is set to 1,000
Set to times. However, the value of the determination value NEa is not limited thereto.

In general, it is difficult to unambiguously determine how many evaluations should be set. This is because the degree of convergence of the search depends on the type and number of control parameters, the contents of the initial set, the method of generating a new set, the conditions for additional registration,
This is because it greatly depends on search conditions such as

In order to obtain a large number of excellent sets exhibiting even slightly superior group management performance, the search end determination value NEa may be set to a value as large as possible. However, the cumulative value of the number of searches
When the NE becomes large to some extent, it takes a long time to complete the search, which causes a decrease in search efficiency. Therefore, in order to efficiently obtain a good set, the search end determination value NEa
Must be set as appropriate according to the search conditions.

Returning to FIG. 11, the optimal set extraction program of step 38 corresponds to the extracting unit 20 of FIG. 1, and extracts one optimal set from a plurality of sets. This will be described with reference to FIG.

In FIG. 20, first, in step 381, the average waiting times AWT (1) to AWT (P) are extracted from the group management performance data PRE (1) to PRE (P), respectively, and these are used for optimal value determination. Substituted into the performance evaluation values VPS (1) to VPS (P). Then, in step 382, an initial setting for detecting the optimum set is performed. That is, the search counter RC is set to 1, the best value BVPE of the performance evaluation value is set to 9,999, and the set number B
Set P to 0.

Next, by repeating the processing of steps 383 to 386,
Optimal set with the best performance evaluation value (set number BP)
Is specified. That is, in step 383, the performance evaluation value VPS (RC) is compared with the highest value BVPE checked so far. If a performance evaluation value VPS (RC) better than the maximum value BVPE is detected, the performance evaluation value VPS (RC) is substituted for the maximum value BVPE in step 384, and the set counter BP is set to the set number BP. Set the value. In step 385, the search counter RC is incremented by one. In step 386, it is determined whether or not the investigation has been completed for all the sets.

In step 387, the optimum set data BPD including the optimum set BPS, the elevator specification data ELSY, the traffic flow specification data TRSY, and the search state data SS is created. That is, the contents of the set having the highest value BVPE are substituted for the optimal set BPS, and the simulation condition data is set for the elevator specification data ELSY and the traffic flow specification data TRSY.
Set the same contents as elevator specification data ELSX and traffic flow specification data TRSX in SIM. In the search state data SS, the value of the evaluation number NE up to that time is set.

Finally, in step 388, the optimal set signal 20a including the optimal set data BPD is output to the group management device 1.

Returning to FIG. 11, as described above, when the search for the optimal set is completed, the process returns to step 26 again until the search permission flag FLAG is reset to “0” by the search end determination program and the search end is determined. Steps 26, 27, and 30 to 38 are repeatedly executed. If there is a change in the contents of the elevator specification data or the traffic flow specification data during the search, a re-search is executed. That is, the search permission flag FLAG is changed to “1”, and the steps from step 31 are executed.

-Advantages of Embodiment 1- As described above, according to Embodiment 1, an excellent set can be efficiently generated, and an optimum set can be efficiently searched. Further, since the group management simulation device 2 separate from the group management device 1 is used, the optimum set can be searched without hindering the original group management operation.

In the first embodiment, since a new set can be generated by two types of generation methods, “crossover” and “mutation”, the characteristics of both can be fully exhibited. That is, with respect to the new set to be created,
With appropriate convergence and simultaneous search, the optimal set can be searched early by combining the global search with the local search.

Further, in the first embodiment, the selection probability is weighted based on the magnitude of the performance evaluation value for each parent set, and then the parent set is selected. Therefore, the probability of selecting an excellent parent is reduced. Can be increased, in other words, the probability of generating a good new set that inherits the good qualities of the parent can be increased.

Further, in the first embodiment, the additional reference value is corrected based on the worst value among the plurality of performance evaluation values so that the additional reference value gradually becomes stricter. Processing can be avoided. In addition,
If the end of the search, the correction of the parent set selection condition, and the correction of the parameter selection condition are performed based on the number of additional registrations, appropriate processing corresponding to the degree of progress of the search can be realized.

Further, in the first embodiment, the number of registered sets can be suppressed to a constant value by the deletion processing, so that a reasonable set can be registered in consideration of the storage capacity. As a result, a better set can be selected from as many sets as possible.

Further, in the first embodiment, since the set with the lowest performance evaluation value is deleted in order, only the set with the highest performance evaluation value can be left, and the superior set is always set as the parent set when a new set is generated. Can be.

In the first embodiment, the extraction unit searches for the optimal set at each time point even during the search, and outputs the optimal set. Therefore, the group management device 1 can obtain the optimum set even during the search without waiting for the completion of the search, and can use it. Furthermore, since the search state data (evaluation count NE) added to the optimal set can be obtained during the search, the utility value of the optimal set output during the search can be accurately determined.

In the first embodiment, the search is continued until the number of searches reaches the predetermined number, so that it is possible to prevent the search from being completed before the search is sufficiently performed.

Further, since the first embodiment has a re-search function, even after the search is completed, the elevator specification data or
When any of the traffic flow specification data changes, the search for a good set is automatically restarted. Therefore, even if the command to start the search from the group management device is delayed for some reason, the search can be started early. Therefore, an optimal set corresponding to the latest group management condition can be obtained at an early stage. Further, the re-search function enables the search to be automatically restarted from the beginning under new group management conditions when the elevator specification data or the traffic flow specification data changes even during the search.

In the first embodiment, an initial set group corresponding to each traffic flow specification is prepared in advance. Then, when the search is started, an initial set group that best matches the traffic flow at that time can be selected and initialized. Therefore, a set that is somewhat excellent from the beginning can be set as a parent set, and a quick search can be achieved. When using the set output as the optimal set during the search,
Even in the early stage of the search, a somewhat good group management performance can be obtained.

In the first embodiment, the group management performance data PRF obtained by the simulation device 2 is stored in the storage unit 11. here,
The group management performance data PRF is composed of a plurality of data as shown in FIG. And the performance evaluation value for additional registration
Performance evaluation values VPE (1) to VPE for setting VPN and reference values
(P), performance evaluation values VPD (1) to VPD for deletion determination
(P), performance evaluation values VPS (1) to VPS for optimal value determination
Any data included in the group management performance data PRF is substituted for (P). Therefore, it is not necessary to perform a simulation every time it becomes necessary to acquire each performance evaluation value. Of course, each performance evaluation value is common data (for example,
In the case of using the average waiting time (AWT), only the common data may be stored as the group management performance data PRF.

[General theory of generation method and selection method] There are roughly two methods for generating a new individual (child) of the (n + 1) th generation from a population (parent) of the n generations. First
Is a method of using the newly generated individual (child) as a parent for generating the next new individual (child) of the same generation (n + 1 generation). On the other hand, the second method Is a method that is not used.

That is, Gn (Mn): the population of the nth generation whose population size is Mn Gn * (j): the population size is i and at least the population
A new individual group gn (i) consisting of j new individuals generated based on Gn (Mn): gn (i) an i-th individual of the nth generation gn * (j): a new individual group Gn * ( j), the (j + 1) th new individual gn * (j +
There are the following two methods for generating and adding 1) to generate a new population Gn * (j + 1).

[Generation Method A]: A new individual gn * (j + 1) is used by using only the current generation individual Gn (Mn) as a parent and by crossover or mutation.
[Generation method B]: current generation population Gn (Mn) and new population Gn
A method in which all or part of * (j) is used as a parent and a new individual gn * (j + 1) is generated as a child by crossover or mutation, where Gn (Mn) = ｛gn (1), gn (2) , ..., gn (Mn)｝ Gn * (j) = ｛gn * (1), gn * (2), ..., gn * (j)｝ Gn * (j + 1) = ｛gn * (1), gn * (2),... Gn * (j), gn * (j + 1)} As a modification of the above [generation method B], a new population Gn
* When a part of (j) is set as a parent, the method is limited to only those having a qualification suitable as a parent (hereinafter referred to as [Generation Method
Ba]).

On the other hand, the method for selecting a next-generation population can be classified according to whether or not the current-generation individual is left as the next-generation individual. That is, Gn * (Mn *): a new population having a population size of Mn * Gn + 1 (Mn + 1): a next-generation population having a population size of Mn + 1 [Selection method A]: New population Gn * ( Mn *)
On a predetermined basis, Mn + 1 new individuals gn + 1 (i) (i = 1,
…, Mn + 1) to select the next generation population Gn + 1 (Mn +
Method 1) (Current generation population Gn (Mn) is not left as a next generation individual at all) [Selection method B]: Current generation population Gn (Mn) and new population Gn
* (Mn *) from all or a part of the sample, Mn +
A method in which one new individual gn + 1 (i) (i = 1,..., Mn + 1) is selected and used as the next-generation individual Gn + 1 (Mn + 1), where Gn * (Mn *) = ｛gn * (1) , gn * (2),... gn * (Mn *)｝ Gn + 1 (Mn + 1) = {gn + 1 (1), gn + 1 (2),... gn + 1 (Mn + 1)} , Current generation population
Among the Gn (Mn), there is also a method (hereinafter referred to as [selection method Ba]) in which an individual who cannot be a parent of the new population Gn * (Mn *) is not left as a next-generation individual. In addition, a new population Gn * (Mn
*) In the method of not leaving new individuals who do not qualify as parents as next-generation individuals (hereinafter referred to as [Selection method B]
b]).

[Example of Combination of Generation Method and Selection Method] (A) When implementing an optimal set search method combining the above [Generation Method B] and [Selection Method B], the following is performed.

That is, the storage unit 11 stores the current generation set group area,
It is divided into two areas: a set group area to be newly registered.
Then, the generation unit 12 generates a new set group using the current generation set group and the additionally registered set group.
Then, the adding unit 15 selects a new set from the new set group based on a predetermined standard (including a case where all new sets are selected unconditionally), and performs additional registration. On the other hand, the deletion unit 16 sets the number of additional registrations to a predetermined value (for example,
Each time (Ps−Pe + 1)), a certain number of sets are selected from the current generation set group and the additionally registered set group based on a predetermined standard, and the set group is set again as the current generation set group. I do. And these are repeated.

Therefore, in the case of such a combination, the generation unit 12 and the addition unit 15 have the function of [generation method B], and the addition unit 15 and the deletion unit 16 have the function of [selection method B].

In the first embodiment, [generation method Ba] which is a modification of [generation method B] and [selection method Bb] which is a modification of [selection method B] are used.

Focusing on the fact that the adding unit 15 of the first embodiment additionally registers only the new set having a suitable qualification as a parent and uses it as one of the parents of the next new set. It can be said that 15 plays a part of the function of [generation method Ba] together with the generation unit 12. At the same time, paying attention to the fact that the adding unit 15 of the first embodiment additionally registers only those having a qualification suitable as a parent and makes it a candidate for a next-generation parent set. Sorting method B
It can be said that it plays a part of the function of b).

(B) When combining [Generation Method A] and [Selection Method A], the following is performed. That is, the storage unit 11 is divided into two areas: a set group area of the current generation and a set group area to be additionally registered. Then, the generation unit 12 generates a new set group from the current generation excellent set groups. Then, by the adding unit 15, from the new set group,
A plurality of sets are selected based on a predetermined standard and additionally registered. On the other hand, the deletion unit 16 sets the number of additional registrations to a predetermined value (for example, Pe
), The set group of the current generation is all deleted, the set group additionally registered therein is moved in its entirety, the generation is updated, and these are repeated.

Therefore, in the case of such a combination, the generation unit 12 of the first embodiment has the function of [generation method A], and the addition unit 15 and the deletion unit 16 have the function of [selection method A].

(C) When combining [Generation Method A] and [Selection Method B (or Bb)], the following is performed.

That is, the storage unit 11 is divided into an area for a set group of the current generation and an area for a set group to be additionally registered, and the generation unit 12 generates a new set group from the set group of the current generation. . Then, the adding unit 15 selects a new set group from among the new set groups based on a predetermined criterion and additionally registers. On the other hand, every time the number of additional registrations reaches a predetermined value (for example, (Ps−Pe + 1)), the deletion unit 16 sets a set based on a predetermined criterion from the current generation excellent set group and the additionally registered excellent set group. A group is selected, and the set group is set as an excellent set group of the current generation. And these are repeated.

Therefore, in the case of such a combination, the generation unit 12 of the first embodiment has the function of [generation method A], and the addition unit 15 and the deletion unit 16 have the function of [selection method B (or Bb)]. To have.

(D) When combining [Generation method B (especially Ba)] and [Selection method A], the following is performed.

That is, the storage unit 11 stores an area for the current generation set group,
The area is divided into two areas for a set group to be newly additionally registered.
Then, the generation unit 12 generates a new set group from the current generation excellent set group and the additionally registered excellent set group. The adding unit 15 selects a new excellent set group from the new set group based on a predetermined criterion, and performs additional registration. On the other hand, the deletion unit 16 sets the number of additional registrations to a predetermined value (for example, Pe
Each time), delete all the excellent sets of the current generation,
The set group additionally registered there is moved in its entirety. And these are repeated.

Therefore, in the case of such a combination, the generation unit 12 and the addition unit 15 of the first embodiment have the function of [generation method B (especially Ba)], and the addition unit 15 and the deletion unit 16 have the [selection method A]. ] Function.

[Regarding Performance Evaluation Values] For reference, the following summarizes the means of use and the purpose of use for each performance evaluation value used in the above examples.

[19] Performance evaluation value for determining the optimal set: VPS (1) to VPS
(P)-Means of use: extractor 20-Purpose of use: used to select the optimal set [20] Performance evaluation value for additional registration determination: VPN-Means of use: adder 15-Purpose of use: addition of new set Used for registration judgment [21] Performance evaluation value for deletion judgment: VPD (1) to VPD (P) ・ Usage means: Deletion unit 16 ・ Usage purpose: To judge deletion of registered set Use [22] Performance evaluation value for setting additional reference value: VPE (1) to VPE
(P)-Means of use: reference value correction unit 18-Purpose of use: used as a reference when correcting additional reference value BX [23] Performance evaluation value for setting parent selection probability: VPS (1) to VPS
(P)-Means of use: generation unit 12-Purpose of use: used to select parent set Here, the superiority of each set can be determined from various angles. For example, it can be determined from the viewpoint of how much the control target required by the group management device can be satisfied. In this case, the group management performance value (see FIG. 7) directly related to the requested control target can be used as it is as the performance evaluation value.

Regarding crossover, the superiority of each set as a parent is
The determination can be made from the viewpoint of the diversity in the storage unit. This is because crossing between sets having different characteristics can increase the probability of generating a better set.

As the performance evaluation value indicating the diversity, a “distribution index” can be used. The distribution index can be defined as, for example, the number of other sets whose distance between sets is within a predetermined value around each set. Here, the inter-set distance is defined in a multidimensional space defined by a plurality of set components. The distribution index indicates the similarity between sets, and a set having a smaller value has higher performance as a parent.

The distribution index can be defined as the sum of distances from other sets. In this case, the higher the value, the higher the performance as a parent. Further, the distribution index can be defined as the number of other sets separated by a predetermined distance or more.
In that case, the higher the value, the higher the performance as a parent.

 Next, a method of obtaining the performance evaluation value will be described in detail.

In the first embodiment, all the performance evaluation values are configured by the average waiting time AWT. However, the content of each performance evaluation value can be different depending on the purpose of use.
For example, each performance evaluation value E may be calculated using different performance evaluation functions.

Generally, a performance evaluation function related to group management performance is expressed as follows. Note that F (X) represents a function of X.

E = F (X1, X2, ..., Xi, ..., Xn, T1, T2, ..., Ti, ..., Tn) ...
[24] where, n: number of evaluation items of group management performance Xi: performance value in evaluation item i (i = 1, 2,...,
n) Ti: performance reference value of evaluation item i (i = 1, 2,..., n) Here, the performance reference value Ti represents a “target value” that should ultimately be reached as group management performance, or In some cases, the “limit value” must be satisfied at least.The “limit value” includes an “upper limit value” and a “lower limit value”. Whether the performance reference value is given as a “target value”, “upper limit value” or “lower limit value” differs depending on where the purpose of the group management control is set.

In the performance evaluation function [24], when the performance reference value Ti means “target value”, the smaller | Xi−Ti |, the better the performance. Is, the larger the (Ti-Xi), the better the performance, and the performance reference value Ti
Means the “lower limit”, the higher the (Xi−Ti), the better the performance.

In short, the performance evaluation values shown in the above [19] to [23] are determined by the performance evaluation function, the evaluation item and the performance reference value therein.

Hereinafter, the calculation method of the performance evaluation value will be described with some specific examples.

[25] Performance evaluation function example 1 (in the case of the first embodiment) • Control purpose: “Make the average waiting time as small as possible” • Performance reference value: T1 = 0 seconds (T1: “Target value” of the average waiting time) -Performance evaluation function: E = | AWT-T1 | = AWT (AWT: average waiting time)-Judgment of additional registration: E <BX (BX: additional reference value, for example, BX
(= 15 seconds) [26] Example 2 of performance evaluation function (in the case of Example 2 described later)-Control purpose: "Make the average waiting time as close as possible to a predetermined value"-Performance reference value: T1 = 20 seconds (T1: average "Target value of waiting time" ・ Performance evaluation function: E = | AWT-T1 | (AWT: average waiting time) ・ Judgment of additional registration: E <BX (BX: additional reference value, for example, BX
[27] Performance evaluation function example 3-Control purpose: "Make the average waiting time AWT, long waiting rate RLW, and forecast miss rate RPE as close as possible to their respective target values."-Performance reference value: T1 = 15 Seconds, T2 = 2%, T3 = 3%. (However, T1: "target value" of average waiting time, T2: "target value" of long wait rate, T3: "target value" of missed forecast rate)-Performance evaluation function: E = Ap × | AWT-T1 | + Aq × | RLW−T2 | + Ar × | RPE−T3 | (However, Ap, Aq, and Ar are weighting factors) • Judgment of additional registration: E <BY (BY: Overall evaluation standard value, for example, BY = 10) [28] Example 4 of performance evaluation function ・ Control purpose: “To reduce the average waiting time AWT, long waiting rate RLW, and missed forecast rate RPE comprehensively as much as possible” ・ Performance reference values: T1 = 0 seconds, T2 = 0%, T3 = 0%.
(However, T1: "target value" of average waiting time, T2: "target value" of long wait rate, T3: "target value" of missed forecast rate)-Performance evaluation function: E = Ap x AWT + Aq x RLW + Ar x RPE (However, Ap, Aq, and Ar are weighting factors.) Judgment of additional registration: E <BY (BY: Overall evaluation reference value; for example, BY = 1000) [29] Example 5 of performance evaluation function • Control purpose: " The average waiting time AWT, long waiting rate RLW, and missed forecast rate RPE should be as large as possible in the respective allowable ranges. ”-Performance standard values: T1a = 15 seconds, T1b = 3 seconds, T2 = 2 %, T3 =
3%. (However, T1a: “target value” of average waiting time, T1b: allowable range of average waiting time deviation, T2: “upper limit value” of long waiting ratio, T3:
(Upper limit of forecast deviation rate) ・ Performance evaluation function: E = f (| AWT−T1a | −T1b) + f (RLW−T2) + f (RPE−T)
3) (where f (X) is f (X) = 1 and X <X when X ≧ 0.
When 0, represents a function where f (X) = 0) ・ Additional registration judgment: E <BY (BY: Overall evaluation reference value, for example, BY = 1) [30] Example 6 of performance evaluation function ・ Control purpose: "The average waiting time AWT falls within a predetermined tolerance from the best (minimum) BVPE among them,
And make the long wait rate RLW as small as possible. ”-Performance standard values: T1a = BVPE (seconds), T1b = 2 seconds, T2 =
0%. (However, T1a: "target value" of average waiting time, T1b: allowable range of average waiting time deviation, T2: "target value" of long waiting ratio)-Performance evaluation function: E = (100-RLW) x f ( T1b− | AWT−BVPE |) (where f (X) is f (X) = 1 and X <X when X ≧ 0)
(When 0, represents a function that satisfies f (X) = 0.) • Determination of optimal value: Max {E} In the above [25] performance evaluation function example 1 and [26] performance evaluation function example 2, the evaluation items are average. This is the case of one waiting time, and since the control purpose is simple, the evaluation function can be configured relatively easily. Therefore, instead of the average waiting time, FIG.
(Of course, evaluation items other than those shown in FIG. 7) are very easy to use.

On the other hand, when there are a plurality of evaluation items, the control purposes also become diversified, so that the configuration of the performance evaluation function becomes complicated. In the case of the above [27] Performance Evaluation Function Example 3 and [28] Performance Evaluation Function Example 4, the performance evaluation function was constituted by the weighted sum of the deviation between the evaluation item and the target value. As described above, a method of comprehensively evaluating the degree of attainment of a goal in consideration of the priority of each evaluation item is generally used. In particular, this is a very convenient method when an evaluation item having a conflicting control target is included therein.

Also, as shown in [29] Performance Evaluation Function Example 5, for each evaluation item, within a set allowable range (for example, below the upper limit, above the lower limit, or the deviation from the target value is within a predetermined value). , Etc.), the number of evaluation items in which the group management performance value has fallen is determined, and the group management performance can be evaluated based on the number.

In addition, it is possible to calculate an evaluation value by using two or more different performance evaluation functions, and to perform addition registration determination, deletion determination, optimal set determination, and the like based on a combination of the plurality of evaluation values.

Also, as shown in [30] Performance evaluation function example 6, the average waiting time is within a predetermined allowable range from the minimum value (E1 = | AWT−
BVPE |, E1 ≦ T1b), and when the long wait ratio is the minimum (E2 = RLW, Min {E2}), first, the performance evaluation function E1 , An optimal value candidate can be selected, and then the optimal value can be finally selected by the performance evaluation function E2. Also, as shown in [30] Example 6 of performance evaluation function, two performance evaluation functions and determination conditions are combined and written into one performance evaluation function and determination condition, and an optimal set is determined by the performance evaluation function and determination conditions. You can choose.

In the first embodiment, the selection probability of the parent set is
It was set based on the performance evaluation value. However, the present invention is not limited to this, and the selection probabilities can be equalized. In that case,
A new set having similar characteristics is likely to be generated, and thus the diversity of the set in the storage unit 11 may be lost. When diversity is lost, there is a problem of converging to a local solution from the beginning of the search. Therefore, when it is necessary to avoid the problem of the initial convergence, or when it is necessary to reduce the calculation amount, the selection probability of the parent set can be set equally.

Second Embodiment Next, a second embodiment will be described with reference to FIGS. In the description of the second embodiment, a description will be given mainly of portions different from the first embodiment.

FIG. 21 is a diagram corresponding to FIG. 1 and shows the overall configuration of the second embodiment. In FIG. 21, the performance reference value setting device 3
It is constituted by a personal computer and outputs a reference value signal 3a to the group management device 1. Here, the reference value signal 3a
Consists of a “performance reference value” for group management performance and a “control reference value” for controlling the search device. In this embodiment, the performance reference value is a “target value” of the average waiting time, and the control reference value is a “designated value” given to the additional reference value BX.
It is. The search device 10 (evaluation means 13, reference value updating means 18, re-searching means 19, initial setting means 21) sends the reference value signal 3a
May be directly input.

FIG. 22 corresponds to FIG. 3 of the first embodiment, and FIG.
Is stored. In FIG. 22, TGT is one of the data that constitutes the content of the search condition signal 1a, and data TAW (waiting time target value) representing the “target value” of the average waiting time AWT.
And data TCB (additional reference specified value) representing the specified value to the additional reference value BX. For example, the waiting time target value TA
W is set to 5 seconds, and the additional reference designation value TCB is set to 3 seconds. TAWX is data in which the input waiting time target value TAW is transcribed, and is used to calculate a performance evaluation value.

When the reference value signal 3a is input, the group management device 1 sets the performance reference value (the waiting time target value) included in the reference value signal 3a.
TAW) and the control reference value (additional reference specification value TCB). Then, similarly to the operation of step 234 in FIG. 10, the data is constituted by elevator specification data ELS, traffic flow specification data TRS, search command data SCM, waiting time target value TAW, additional reference designation value TCB, and search condition signal 1a. The search condition signal 1a is output to the search device 10.

When the search is started in the search device 10 (see FIG. 11), generation, evaluation, addition, deletion, and addition reference value correction are performed in order, as in the first embodiment. In addition, based on the search start determination program 26 (see FIG. 25), the search condition signal 1a
Is input, as shown in FIG. 22, elevator specification data ELS, traffic flow specification data TRS, search command data SCM,
The waiting time target value TAW and the additional reference specified value TCB are stored in RAM10C.
Is stored.

The second embodiment is characterized in the contents of the addition program 34 and the deletion program 35.

In the second embodiment, the above [26] Example 2 of performance evaluation function
The performance evaluation value E is calculated based on the performance evaluation function (E = | AWT-T1 |) shown by the following equation, and the performance evaluation value E is compared with the additional reference value BX. Then, when the performance evaluation value E is smaller than the additional reference value BX, additional registration is determined. That is, the difference between the actual value and the target value is compared with the specified value, and when the difference is smaller than the specified value, additional registration of a new set is performed. Therefore, in step 341 of the additional program 34 (see FIG. 16), the operation [VPN ← | AWT−TAWX |] is performed, and the performance evaluation value VPN for additional registration determination is set.

Step 352 of the deletion program 35 (see FIG. 17)
, An operation [VPD (P) ← | AWT (P) −TAWX |] is performed, and the performance evaluation values VPD (1) to VPD
(P) is set.

FIG. 23 shows an additional reference value correction program in the second embodiment.
36 contents are shown.

In FIG. 23, first, in step 401, the additional reference designation value T
CB is read, and is substituted for the correction value CBX. Then, in the next step 402, the additional reference value BX is rewritten with the correction value CBX. In step 403, the waiting time target value TAW is read, and the currently used performance reference value (waiting time target value TAWX) is rewritten based on this value.

When the correction of the additional reference value is completed in this manner, the search end determination and the optimal set extraction processing are performed as in the first embodiment (see FIG. 11). However, in the optimum set extraction program 38, the method of calculating the performance evaluation value is different from that of the first embodiment. That is, in step 381 of the optimal set extraction program 38 (see FIG. 20), the performance evaluation values VPS (1) to VPS (P) for determining the optimal set are calculated as [VPS (1) ← | AWT
(1) −TAWX |,..., VPS (P) ← | AWT (P) −TAWX |].

After the processing of the optimum set extraction program 38, the process proceeds to the processing of the re-search determination program 25. The operation procedure of the search start determination program 26 in the re-search determination program 25 will be described with reference to FIG. FIG. 24 is a diagram of the first embodiment.
Equivalent to 12.

In FIG. 24, in step 261, the search condition signal 1a is input from the group management device 1, and the elevator specification data ELS,
The traffic flow specification data TRS, the search command data SCM, the waiting time target value TAW, and the additional reference designation value TCB are stored in the RAM 10C. Then, in steps 262 to 264, similarly to the search start determination program 26 of the first embodiment (see FIG. 12), the search command data
A change in M from “0” to “1”, a change in the elevator specification ELS, or a change in the traffic flow specification TRS is detected. In step 265, if there is at least one change, the search start flag STR is set to "1", and a command is issued to start the search after initial setting of mode 1 described later.

On the other hand, search command data SCM and elevator specification data EL
If there is no change in either S or the traffic flow specification data TRS, it is determined in step 267 whether or not the waiting time target value TAW has changed. That is, the waiting time target value TAW is compared with the currently set waiting time target value TAWX to determine whether there is any change. If TAW ≠ TAWX, “changed”
Is determined, and in step 265, the search start flag STR is set to “1”.
Set to. On the other hand, if TAW = TAWX, it is determined that there is no change, and step 269 is executed.

In step 269, it is determined whether or not the additional reference specified value TCB has changed. That is, the additional reference designation value TCB is compared with the currently set additional reference value BX to determine whether there is any change. If BX = TCB, it is determined that there is “no change”, and in step 266, the search start flag STR is set to “0”.
If the search is in the end state, the end state is maintained. On the other hand, if BX ≠ TCB, it is determined that “change has occurred”, and in step 268, the search start flag STR is set to “2”, and a command is issued to start the search after initializing the mode 2.

Here, the initial setting of the mode 1 refers to a process of performing an initial setting in the storage unit and a general initial setting (setting of initial values such as the number of evaluations NE and the number of additional registrations NR). Further, the initial setting of mode 2 refers to a process of performing only general initial setting without performing initial setting in the storage unit.

In the above step 269, when the additional reference value BX is changed, the initial setting of the mode 2 is performed when the additional reference value BX is changed.
Is changed to a stricter (ie, smaller) value than before, there is no problem if the search is advanced on an extension of the previous search. Rather, in terms of convergence,
The search efficiency is better if the search generated and selected up to that point is used as the initial set and the search is restarted. In particular, when the performance evaluation function includes only a single evaluation item and only the additional reference value BX is changed, the initial setting of the mode 2 is suitable.

Now, as described above, as a result of determining whether to keep the current search (continuous state or end state) for the optimum set or to start over from the beginning, the search start flag STR = "0" Is set. In this case, in FIG. 11, the process proceeds from step 27 to step 30, where a search permission flag FLAG
Is determined by the value of If the search is in the middle of the search, FLAG = “1”, so the process proceeds to step 31 again, and the search end determination program 37 resets the search permission flag FLAG to “0”.
The procedure of steps 31 to 38, steps 26 and 27 is repeated until it is determined in step 27 that the search has been completed.

When the search is completed, a search end determination program
Since the search permission flag FLAG is set to "1" by 37, a re-search is waited by repeating the procedure of steps 26 → 27 → 30 → 26.

If during or after the search, the elevator specification data ELS, traffic flow specification data TRS, performance reference value TAW,
If the control reference value TCB has been changed, the search start determination program 26 sets the search start flag STR to "1" or "2", and the process proceeds from step 27 to the initial setting program 28.

FIG. 25 shows the initial setting program 28. FIG. 25 corresponds to FIG. 13 of the first embodiment.

In FIG. 25, in step 284, it is determined whether or not the storage unit needs to be initialized according to the value of the search start flag STR. If the initial setting of the mode 1 has been designated, that is, if the search start flag STR is “1”, similarly to the initial setting program 28 of the first embodiment (see FIG. 13), step 28 is executed.
In step 1, an initial set group and group management performance data corresponding to the traffic flow specification data TRS are read. Then, step 282
The initial setting is performed using the initial set group and the group management performance data.

Then, in step 285, the additional reference value correction program is started. The additional reference value correction program in step 285 has the same contents as the program 36 shown in FIG. In this step 285, the new additional reference value (BX)
And a performance reference value (a waiting time target value TAWX) are set. Then, similar to the initial setting program 28 (see FIG. 13) of the first embodiment, general initial settings are performed in step 283. The above is the initial setting of the mode 1.

On the other hand, if the search start flag STR is "2" in step 284, that is, if the initial setting of mode 2 is designated, only the general initial setting of step 283 is executed.
That is, the processing of steps 281 and 282 is not performed, and the set group registered in the storage unit before the start of the current search is used as the initial set group. That is, at the end of the previous search, the excellent set EPS (1) remaining in the storage unit 11
~ EPS (P) and its group management performance data PRE (1) ~ PRE
(P) is used. The above is the initial setting of the mode 2.

As described above, in the second embodiment, the target value TAWX of the average waiting time and the designated value BX of the additional reference value can be externally input by the performance reference value set value 3. Therefore, it is possible to search for an optimal set that matches a desired group management control policy or search policy.

Further, in the second embodiment, a change in the reference value TGT (a waiting time target value TAW, an additional reference value BX) can be detected during the search or after the search is completed, and the search can be performed again. Therefore, when the control policy of the group management is artificially changed or when the reference value TGT is artificially changed, the re-search can be automatically performed, and the optimal set can be quickly obtained.

In the second embodiment, the initial setting mode of the mode 1 and the mode 2
Can be selected, so that an appropriate initial setting can be made according to the situation at the start of the search. For example, when the additional reference value BX is finely modified, the re-search can be terminated early by the initial setting of the mode 2.
Of course, when the change amount of the additional reference value BX is large, it is desirable to apply the initial setting of the mode 1. Conversely, in the second embodiment, when a change in the performance evaluation function (evaluation item, performance reference value, configuration) is detected, the initial setting of mode 1 is applied. In
The initial setting of mode 2 can also be applied.

In short, the mode selection problem of whether to apply the initial setting of mode 1 or the initial setting of mode 2 is as follows.
Under the new search condition, the problem is as to which of the set group obtained up to that point and the initial set group GPS0 can realize an efficient search. However, the determination requires a great deal of calculation time and calculation amount, and is not a practical method. So, as mentioned above,
It is desirable to select a mode according to the change item and the change amount.

The above performance evaluation function is an example, and [24] Performance evaluation function and performance evaluation function examples 1 to 6 ([25] to [30])
Etc. can also be used.

Further, the performance evaluation values are not limited to the performance evaluation values for the additional registration determination described above, but are [19] a performance evaluation value for the optimal set determination, [21] a performance evaluation value for the deletion determination, and [22] a reference. A performance evaluation value for value determination may be used.

Third Embodiment (Another Embodiment of Deletion Unit) Next, another embodiment of the deletion unit 16 will be described with reference to FIG. The following description focuses on the differences from the first embodiment.

FIG. 26 shows the deletion program 35 of the third embodiment. This flowchart corresponds to FIG. 16 of the first embodiment.

First, in step 411, the equation NRH = NR-NRX is executed to calculate the number of new registrations NRH. The number of new registrations N
RH is the number of times a new set has been additionally registered since the previous deletion processing was performed. The number of registrations NRX at the time of the previous determination is a value indicated by the number of additional registrations NR when the previous deletion processing was executed.

Next, in step 412, the number of new registrations NRH is compared with the deletion start determination value NRa to determine whether it is time to delete. For example, the deletion start determination value NRa is set to 10 times.

If NRH <NRa, it is determined that it is not time to delete the set, and the program immediately exits the program 35. On the other hand, if NRH ≧ NRa, it is determined that it is time to delete, and in step 413, the value of the number of additional registrations NR at the present time is set to the number of registrations NRX at the previous determination, and updating is performed.

Then, steps 414 to 422 are repeated, and the number of units P
The set is deleted until becomes the deletion end determination value Pe. The initial value of NRX is initially set to 0 (not shown) in step 283 of the initial setting program 28 (see FIG. 13).

In step 414, the distance DST (i, j) between the sets is calculated (where i, j = 1, 2,..., P, i 、 j).

This distance is DST (i, j), two sets as follows
Norm for i, j.

DST (i, j) = {EPU (i) -EPU (j)} ... [31] where i, j = 1,2, ..., P, i ≠ j Here, each set EPS (i) (i = 1, 2,..., P) are normalized for each parameter. That is, each parameter value is a percentage from 0 as the maximum value that each parameter can take.
Expressed as any value between 100.

For example, if the maximum value that can be taken by the packed evaluation coefficient Ca, which is one of the parameters, is 50,000, the set EPS in FIG.
Regarding (1), the value obtained by normalizing the parameter value of the packed evaluation coefficient Ca is calculated as (10,000 ÷ 50,000) × 100 = 20. Similarly, when the parameter value of the forecast deviation coefficient Cb (maximum value: 1,600) is normalized, (400 ÷ 1,600) × 100
= 25 is calculated.

Therefore, since there are a total of 25 control parameters, 0
≦ DST (i, j) ≦ 500. The normalized parameter values are stored in the optimal set extraction program 38 (see FIG. 20).
In step 387, when the optimal set data BPD is created, it is converted back to the original value (for example, the fullness evaluation coefficient Ca
In the case of, it is re-converted to 20 x 50,000 = 10,000).

Next, in step 415, a set pair Pd1 and Pd2 that minimizes the distance DST (i, j) is selected. Then, in step 416, it is determined whether or not the two sets Pd1 and Pd2 have similar characteristics. That is, the distance DST (Pd
1, Pd2) and the determination value DSTa are compared, and similarity is determined from the result. As the determination value DSTa, for example, 25 is set.

If DST (Pd1, Pd2) ≦ DSTa, step 417
Then, the average waiting times AWT (Pd1) and AWT (Pd2) are extracted from the group management performance data PRE (Pd1) and PRE (Pd2) of the two sets Pd1 and Pd2, respectively, and are deleted. Are set as the performance evaluation values VPD1 and VPD2.

In step 418, the performance evaluation values VPD1 and VPD2 are compared to determine a set to be deleted. If VPD1 <VPD2, it is determined that the set number Pd2 is to be deleted, and the process proceeds to step 419. Then, the registration of the set EPS (Pd2) and the group management performance data PRE (Pd2) is deleted. further,
The value of the number of registered units P is also reduced by one. The set number is re-assigned to the remaining sets, and the process of step 419 ends.

On the other hand, in step 418, if VPD1 ≧ VPD2, it is determined that the item to be deleted is set number Pd1,
Proceed to step 420. Then, the registration of the set EPS (Pd1) and the group management performance data PRE (Pd1) is deleted. Further, the value of the number P of registered units is also reduced by one. Then, the set numbers are re-assigned to the remaining sets, and the process of step 420 is terminated.

If in step 416, DST (Pd1, Pd2)> DSTa
If so, it is determined that the characteristics are not similar, and step 42
The process proceeds to step 1, where the performance evaluation values VPD (1) to VPD (P) for deletion determination for all the unit numbers 1 to P are obtained, and a set having the worst value is specified from these. Set the set number as Pd1. This step 421
Are the same as steps 352 to 357 in the deletion program 35 of the first embodiment (see FIG. 17), and therefore description thereof is omitted. Note that the deletion set number RP in FIG. 17 corresponds to the set number Pd1.

In step 422, it is determined whether or not the number P of sets after deletion has become equal to or less than the deletion end determination value Pe. If not, the processing in steps 414 to 422 described above is repeated, and P ≦ Pe
At this point, the processing of the deletion program 35 ends.

As described above, in the third embodiment, a pair having similar characteristics is specified based on the distance DST between sets, and one set of the pair is deleted. Can be left in the storage unit 11, and diversity in the storage unit 11 can be secured. Further, when a pair is selected, a pair having the highest similarity (closest distance) is preferentially selected, so that a plurality of sets having different characteristics can be left as much as possible.

In the third embodiment, when one of the pair is deleted,
Since the set having the better performance evaluation value is left and the set having the worse performance evaluation value is deleted, the group management performance of the plurality of sets in the storage unit 11 can be maintained at a high level as a whole.

In the third embodiment, when there are no more similar pairs to be deleted, the sets with the worst performance evaluation values are deleted in order. Therefore, unnecessary sets can always be deleted up to the deletion end determination value Pe.

Fourth Embodiment (Another Example of Search End Determination) Another embodiment of the search end determination unit will be described below with reference to FIG. In the description of the fourth embodiment, parts different from the first embodiment will be mainly described.

In step 371 of FIG. 19, the number of additional registrations NR is used instead of the number of evaluations NE, and the search end determination value NRb is used instead of the search end determination value NEa. The number of additional registrations N
As R, for example, 200 times is set.

That is, when NR <NRb, the search permission flag FLAG is set to “1” in step 372 to continue the search, and NR ≧ NR
When it becomes b, the search permission flag FLAG is set to "0" in step 373.
To end the search.

Here, in general, when the number of searches NE is large and the search has converged to some extent, the ratio of the number of additional registrations to the number of searches decreases. Therefore, in order to perform a more efficient search, considering the result of the search process, that is, the number of additionally registered sets,
It is necessary to determine the end of the search. For this reason, in the fourth embodiment, the number of additional registrations indicating the number of additional registrations
The search is continued until the NR reaches the search end determination value NRb.

As described above, according to the fourth embodiment, it is possible to prevent the search from ending before the search is sufficiently performed.

Fifth Embodiment (Another Embodiment of Search End Determination) Next, another embodiment of the search end determination will be described with reference to FIG. In the description of the fifth embodiment, a description will be given focusing on portions different from the first embodiment.

FIG. 27 shows a search end determination program 37 of the fifth embodiment, and corresponds to FIG. 19 of the first embodiment.

In step 431, the elapsed evaluation frequency NEH is calculated by the expression NEH = NE-NEX. Here, the elapsed evaluation number NEH is the number of new evaluations since the previous end determination was made. The number of evaluations NEX at the time of the previous determination indicates the value of the number of evaluations NE at the time of performing the previous end determination. Then, in step 432, it is determined whether or not a predetermined number or more has been evaluated from the previous time. If the elapsed evaluation number NEH is less than the predetermined value NEb, in step 433, the search permission flag FLAG is set to "1" to continue the search. In addition, as the elapsed evaluation number NEH, for example, 20 times is set.

If it is determined in step 432 that the elapsed evaluation times NEH
If yes, go to step 434 where the success indicator RSC
Is calculated. First, the number of new registrations NRH is calculated by the equation NRH = NR-NRX. Here, the number of new registrations NRH
Indicates the number of times a new set has been additionally registered since the previous end determination was made. And the success indicator RSC is R
It is calculated by the formula SC = NRH ÷ NEH. Here, the number of registrations NRX at the previous determination represents the value of the number of additional registrations NR at the time of performing the end determination. In the next step 435, based on the current number of evaluations NE and the number of additional registrations NR,
Update the number of evaluations NEX at the previous determination and the number of registrations NRX at the previous determination.

In step 436, the number of evaluations NE, the search end judgment value NEc,
It is determined whether to end the search based on the success index RSC and the search end determination value RSCa. Here, the search end judgment value
As NEc, for example, 600 times is set. As the search end determination value RSCa, for example, 0.05 is set.

If NE <NEc or RSC ≧ RSCa, it is determined that the search has not been sufficiently performed, and the search permission flag FLAG is set to “1” in step 433 to continue the search. If NE ≧ NEc and RSC <RSCa, it is determined that the search has been sufficiently performed, and in step 437, the search permission flag FLAG is set to “0” to end the search.

In step 436, the condition regarding the number of evaluations NE was included in the initial stage of the search because the success index RSC was low due to the magnitude of the initial set GPS0, the crossover rate CR, and the mutation rate MR. This is to prevent the search from ending without the evaluation of. If there is no such a problem, the condition for the number of evaluations NE is not required as a condition for determining the search end, and the condition for the success index RSC is sufficient.

As described above, in the fifth embodiment, the end of the search is determined based on the success index RSC based on the number of evaluations and the number of additional registrations. Therefore, it is possible to accurately determine whether the search has sufficiently converged. Therefore, it is not necessary to repeat the search needlessly, and an efficient search for the optimal set can be performed.

Further, in the fifth embodiment, it is detected that the search is the initial period based on the number of evaluations NE, and within this initial period, the search is not terminated even if the success index RSC becomes less than the search end determination value RSCa. Therefore, there is no possibility that the search is terminated without performing a sufficient number of evaluations.

Sixth Embodiment (Another Embodiment of Search End Determination) Another embodiment of the search end determination unit 17 will be described with reference to FIG. In the description of the sixth embodiment, the first embodiment will be described.
The following description focuses on the differences from the first embodiment.

FIG. 28 shows a search end determination program 37 of the sixth embodiment, and corresponds to FIG.

In step 451, the distance DST (i, j) between each set is calculated. This distance DST (i, j) is calculated according to the above equation [31]. This calculation is the same as step 414 of the deletion program 35 (see FIG. 26) of the third embodiment.

In step 452, the distance DST obtained as described above
Based on (i, j), the number of similar sets, ie, DST
Number of sets where (i, j) ≤ DSTa NDST (number of similar sets)
Is counted. Note that DSTa is a determination value for determining whether or not the sets are similar, and is set to 25 in the sixth embodiment as in the third embodiment.

In the next step 453, a search end determination value NDSTa is calculated. In general, as the search converges, many sets tend to concentrate around the optimal set. As one of the indicators to grasp this tendency, the number of similar sets ND
Using ST, convergence of the search is determined by the ratio of the number of similar sets NDST to the number of combinations of all sets. Assuming that the determination criterion for the ratio is 80%, when the number of registered sets is P, the number of all combinations is ΔP × (P−
1) Since {2} is set, the search end determination value NDSTa is set to NDSTa
= {P × (P-1) ÷ 2} × 0.8.

Here, when the number of additionally registered sets reaches the maximum value Pmax, unnecessary sets are deleted based on a predetermined criterion. At this time, the number of similar sets NDST changes depending on how the sets are deleted. Similarly, the number of similar units NDST changes depending on the set generation method and the additional registration method. Therefore, the above criterion is not limited to 80% and needs to be appropriately corrected based on other factors.

Thus, in step 454, it is determined whether or not to end the search based on the number of similar units NDST and the search end determination value NDSTa. If NDST <NDSTa, it is determined that the search has not been performed sufficiently, and in step 455, the search permission flag FLAG is set to “1” to continue the search, and the search end determination program 37 The process ends. If it is determined that NDST ≧ NDSTa and the search has been sufficiently performed, the search permission flag FLAG is set to “0” to end the search in step 456, and the processing of the search end determination program 37 ends. .

As described above, in the sixth embodiment, since the end of the search is determined based on the inter-set distance DST, the convergence of the search can be accurately detected. Therefore, it is not necessary to repeat the search needlessly, and an efficient search can be performed.

Note that the search end determination condition is not limited to the above condition, and other conditions may be employed.

Seventh Embodiment (Another Embodiment of Extracting Optimal Set) Next, another embodiment of the extracting unit 20 will be described with reference to FIG. In the description of the seventh embodiment, a description will be given focusing on portions different from the first embodiment.

FIG. 29 shows an optimum set extraction program 38 of the seventh embodiment.
This corresponds to FIG. 20 of the first embodiment.

In step 471, group management performance data PRE (1) to PRE
(P) Average waiting time AWT (1)-AWT
(P) and take them out as the first performance evaluation value VPS1 (1).
To VPS1 (P). In step 472, the long wait ratio RL is selected from the group management performance data PRE (1) to PRE (P).
W (1) to RLW (P) are extracted, and these are set as second performance evaluation values VPS2 (1) to VPS2 (P). In step 473, the minimum value of the first performance evaluation values VPS1 (1) to VPS1 (P) is obtained,
This is set as the best value BVPE.

In step 474, the optimal set B is set based on the performance evaluation value.
Select P. That is, from the storage unit 11, [(VPS1
(I)-Find a plurality of sets i satisfying -BVPE) ≤ BZ]. Further, among them, the set with the second performance evaluation value VPS2 (i) that is the smallest is selected as the optimal set BP. That is, two-stage selection is performed. Note that BZ is a reference value indicating an allowable range from the best value BVPE, and is set to 2 seconds in this embodiment.

Then, similarly to the first embodiment, the optimum set data BPD is created, and in the next step 388, the optimum set data BPD is output to the group management device 1.

As described above, in the seventh embodiment, the two-stage selection is applied to the optimal set extraction. Therefore, when the priority is determined among the two evaluation items, the extraction according to the priority is performed. realizable. Of course, three or more stages of selection can be applied. Note that, in the seventh embodiment, the above [30]
It is equal to the content of the performance evaluation function example 6.

[Eighth Embodiment (Other Method of Obtaining Group Management Performance Value)] In place of the simulation apparatus 2 of the first embodiment,
As described in No. 57-57168, the actual group management device 1
, The group management performance for the new set can be obtained. This will be described with reference to FIGS. 30 and 31.

FIG. 30 shows a system configuration of the eighth embodiment, which corresponds to FIG. 1 of the first embodiment. FIG. 31 is a diagram illustrating the operation of the group management device 1, and corresponds to FIG. 9 of the first embodiment. Hereinafter, an eighth embodiment will be described with a focus on portions different from the first embodiment.

In FIG. 30, when a search condition signal 1a is input from the group management device 1 to the search device 10 and a search command is issued, the search device
10 executes the evaluation program (see FIG. 15) in step 33 of the arithmetic program, creates simulation condition data and outputs it as the simulation condition signal 13a as in the first embodiment.

However, as shown in FIG. 30, in this embodiment, the signal 13a is output to the group management device 1. On the other hand, when receiving the signal 13a, the group management device 1 enters the trial operation mode. This operation will be described with reference to the flowchart in FIG.

In FIG. 31, it is determined in step 491 whether or not trial operation is being performed, and in step 492, it is determined whether or not trial operation is started. The trial operation flag FLG is "0", and
If there is no instruction to start the trial operation in the signal 13a, the normal group management operation is executed according to steps 221 to 229.

On the other hand, when the start of the trial operation mode is detected based on the content of the signal 13a in step 492, the trial operation flag FLG is set to “1” in step 493, and the parameter value set currently used is set in step 494. Evacuate temporarily.
Then, instead, the evaluation set (new set) NPSX included in the signal 13a is written.

Then, similarly to the normal group management operation, Steps 221-2-2
Perform group management operation by 29. In addition, during the trial operation, the trial operation flag FLG is “1”, so that the group management operation is performed by steps 221 to 229 via steps 491 to 495.

In this way, the group management operation is performed for a predetermined period (for example, 1).
(Time), the end of the trial operation is detected in step 495, the trial operation flag FLG is reset to “0” in step 496, and the parameter value set saved in step 497 is restored. Calculate group management performance data PRF (average waiting time, long waiting rate, etc.). Then, in step 498, the group management performance data PRF is output to the search device 10 as the group management performance value signal 2a. Thereafter, returning to the normal state, the group management operation is performed in steps 221 to 229.

As described above, when the group management performance signal 2a is obtained by the trial operation, the search device 10 in FIG.
A performance evaluation value VPN is acquired based on the group management performance value signal 2a, and is compared with the evaluation reference value BX to determine whether or not a new set is additionally registered.

As described above, in the eighth embodiment, since the evaluation of the new set is performed by the actual machine, there is a disadvantage that the time required to obtain the optimum set is long. There is an advantage that it can be configured at low cost.

Example 9 (Other Example of Generating a New Set) In Example 1, the crossover rate CR and the mutation rate MR were fixed. The ninth embodiment is characterized in that CR and MR are corrected according to a search situation.

This embodiment will be described with reference to FIGS. The following description focuses on the differences from the first and second embodiments.

FIG. 32 shows the entire configuration of the ninth embodiment, and FIG.
Is equivalent to In FIG. 32, the crossover rate is calculated by the appearance rate correction unit 4.
The CR and the mutation rate MR (selectivity of each generation method) are modified according to the search situation.

FIG. 33 is a diagram showing an arithmetic program according to the ninth embodiment.
This corresponds to FIG. 11 of the first embodiment. Note that in step 50, FIG.
11 except that an appearance rate correction program corresponding to the function of the appearance rate correction unit 4 is added.
Is the same as

FIG. 34 is a diagram showing the appearance rate correction program.
In step 501, the elapsed evaluation number NEH representing the number of times newly evaluated since the previous end determination was made is calculated by the expression NEH = NE-NEX. Here, the number of evaluations at the previous judgment
In NEX, the same value as the number of evaluations NE at the time of the previous end determination is set. In step 502, it is determined whether the evaluation has been performed a predetermined number of times or more since the last time. If the elapsed evaluation frequency NEH is less than the predetermined value NEb (for example, 20 times), the appearance rate correction program is ended.

On the other hand, in step 502, the elapsed evaluation frequency NEH is set to the predetermined value NEb.
If yes, go to step 503 where the success indicator RSC
Is calculated.

First, the number of new registrations NRH representing the number of times a new set has been additionally registered since the previous end determination was made is given by NRH = NR−
It is calculated by the formula NRX. And the success indicator RSC,
It is calculated by the formula RSC = NRH ÷ NEH. Here, the same value as the number of additional registrations NR at the time of the previous end determination is set to the number of registrations NRX at the previous determination. In the next step 504, the number of evaluations NEX at the previous determination and the number of registrations N at the previous determination
Update RX.

In steps 505 to 510, the number of evaluations NE and the first determination value NEd
1, the second determination value NEd2, the success index RSC, the success rate determination value RSCb,
Based on the success rate judgment value RSCc, the crossover rate CR and the mutation rate MR
To correct. For example, NEd1 = 500, NEd2 = 800, RSCb = 0.
10, RSCb = 0.05.

If NE ≦ NEd1 and RSC ≦ RSCb, that is, if the success index is lower than the expected value in the first half of the search, the currently set crossover rate CR and mutation rate MR are inappropriate. 505 → 507 → 509
And make the crossover rate CR smaller than the current value (for example, 0.001)
And make the mutation rate MR slightly larger (for example, 0.001) than the current value.

In step 509, contrary to the above correction, the crossover rate CR is set to be slightly larger (for example, 0.001) than the current value, and the mutation rate MR is set to be slightly smaller (for example, 0.001) than the current value. You can also. Either method can be basically applied to the breakthrough in the sluggish state. That is, if the crossover rate CR is too large and not working well, reduce the crossover rate CR and
MR should be increased. On the other hand, if the crossover rate CR is too small to be successful, the crossover rate CR may be increased and the mutation rate MR may be reduced.

After all, if the search is sluggish, the ratio of the selection probabilities of the respective generation methods may be changed to overcome the sluggish state.

If NE ≧ NEd2 and RSC ≦ RSCc, that is, if the success index is lower than the expected value at the end of the search, it is determined that the search for the excellent set is almost converging, and step 505 is performed. Going from 506 to 508 to 510, the crossover rate CR is made slightly larger (for example, 0.001) than the current value, and the mutation rate MR is made smaller than the current value (for example, 0.001).
Just make it smaller.

If the above requirements are not satisfied, the currently set crossover rate CR and mutation rate MR are determined to be appropriate,
Without modifying those values, the program for modifying the appearance rate ends.

As described above, in the ninth embodiment, the number of evaluations NE
The progress of the search is determined based on the success index RSC and the success index, and the crossover rate CR and the mutation rate MR can be corrected accordingly. Therefore, as compared with the case where the crossover rate CR and the mutation rate MR are fixedly set, an excellent set can be found earlier, and the search time can be shortened. As a result, search efficiency can be improved.

Further, in the ninth embodiment, especially when the success index is lower than the expected value in the initial stage (first half) of the search,
Since the crossover rate CR is set lower than the current value and the mutation rate MR is set higher than the current value, a set with better group management performance can be generated with emphasis on global search by mutation. Performance can be improved. In addition, it is possible to overcome the sluggish state of search.

Further, in Example 9, when the success index is lower than the expected value at the end of the search (second half), the crossover rate CR is set higher than the current value and the mutation rate MR is set lower than the current value. Therefore, local search can be emphasized and the search can be converged at an early stage. In addition, it is possible to overcome the sluggish state of search.

Embodiment 10 (Another Embodiment of Appearance Rate Correction) Another embodiment of the appearance rate correction unit 4 will be described with reference to FIG. FIG. 35 is a diagram showing an appearance rate correction program, which is obtained by partially correcting the appearance rate correction program (see FIG. 34) of the ninth embodiment.

In FIG. 35, in step 502, a certain number of evaluations NEb
(For example, 40 times). Steps
At 506, it is late in the search (ie, NE ≧ NEd
Determine 2). And finally, at step 510, the crossover rate
Set CR slightly larger than the current value (eg, 0.001) and set the mutation rate MR slightly smaller than the current value (eg, 0.001).
Set only smaller. Therefore, at the end of the search,
As the search progresses, the value of the crossover rate CR gradually increases,
Conversely, the value of the mutation rate MR decreases. The operation other than the above is exactly the same as that of the ninth embodiment.

As described above, in the tenth embodiment, as the search progresses, the generation is switched from the emphasis on mutation to the generation mainly based on crossover.
Diversity of group management performance is ensured, while early convergence of the search can be achieved at the end (second half) of the search. As a result, it is possible to efficiently search for a set having better group management performance.

[Embodiment 11 (Another embodiment of crossover pair selection)] Next, Embodiment 11 will be described with reference to FIGS. 36 to 39. The following description focuses on the differences from the first and second embodiments.

FIG. 36 shows a system configuration of the eleventh embodiment.
In this embodiment, the distance between sets is used as a condition for selecting a pair of sets to be crossed (crossover pair). Further, in this embodiment, it is configured such that the condition regarding the distance between sets (hereinafter, distance condition) can be corrected. The parent selection condition correction unit 5 in FIG. 36 corrects the distance condition according to the search situation.

FIG. 37 is a diagram illustrating an arithmetic program of the eleventh embodiment, and corresponds to FIG. 11 of the first embodiment. In FIG. 37, in step 31, a new set is generated. The operation is as described with reference to FIG. 14 relating to the first embodiment. However, in the eleventh embodiment, the process of selecting the crossover pairs PS1 and PS2 (step 317) is significantly different from the first embodiment. Step 52
, A parent selection condition correction program corresponding to the parent selection condition correction unit 5 (see FIG. 36) is added, and the distance condition described above is corrected. Other configurations are the same as in the first embodiment.

FIG. 38 is a diagram showing the operation of step 317 included in the new set generation program 31 (see FIG. 14). In FIG. 38, in step 317a, the value of the counter RC is initialized to 0. In this embodiment, the counter RC counts the number of times a pair is selected using the distance between sets condition. In addition, the distance between sets is the above [31]
It is calculated by the formula.

In the next step 317b, similarly to the first embodiment, the selection sets are weighted based on the magnitude of the performance evaluation value, and then the two sets PS1 and PS2 to be paired are randomly selected. In the next step 317c, it is determined whether or not the set selection using the distance condition has been performed a predetermined number of times (for example, 10 times) or more.

Here, if the two sets PS1 and PS2 satisfying the distance condition cannot be selected even if Steps 317d to 317h are repeated a predetermined number of times or more, in Step 317b, the selected 2 sets are not selected.
The sets PS1 and PS2 are determined as a pair to be crossed.

On the other hand, when the number of set selections using the distance condition is less than the predetermined number (RC <10), the process proceeds from step 317c to 317d, and first, the counter RC is incremented by one. Then, in step 317e, the two sets selected in step 317e
Calculate the distance DST between PS1 and PS2. That is, DST = {EPS (PS1) −EPS (PS2)} is calculated in the same manner as in equation [31]. In addition, each parameter value is set to 0 as described above.
It has been normalized to a number between 100.

In step 317e, when the distance DST is calculated, step 3
At 17f to 317h, it is determined whether or not the distance condition, which is one of the selection conditions of the cross pair, is satisfied.

If the condition selection flag SELS is set to 1 and the first selection condition is specified as a distance condition,
Proceed to step 317f → 317g, where the distance DST is the first
It is determined whether the value is equal to or greater than the selection reference value DSTb1. DST <DSTb1
In the case of, two sets PS1 and PS2 that do not satisfy the distance condition
Is discarded, the process returns to step 317b, and the same process is repeated from the beginning.

On the other hand, if the condition selection flag SELS is set to 2 and the second selection condition is specified as the distance condition, as described above, at step 317h, whether the distance DST is equal to or less than the second selection reference value DSTb2. If DST> DSTb2, it is determined that the selection condition is not satisfied, and the process returns to step 317b and starts over.

In this way, the operations of steps 317b to 317h are repeated until the determination of the distance condition is performed a predetermined number of times (10 times) or more, or until two sets satisfying the distance condition are found. DST ≧ DSTb1 or DST ≦ DST
When two sets b2 are generated, the two sets are used as normal cross pairs PS1 and PS2, and the process proceeds to the next step 318. The procedure after step 318 is the same as in the first embodiment.

As described above, crossover pairs can be selected using the distance condition. The distance condition is corrected by the selection condition correction program 52 according to the search situation.

FIG. 39 shows a selection condition correction program (step 52).
37) is shown in detail.

In steps 521 to 524, the same processing as in steps 501 to 504 of the appearance rate modification program 50 (see FIG. 34) of the ninth embodiment is performed, and the success index RSC is calculated. This success metric is
Conceptually, it is defined as the number of additional registrations / the number of evaluations.

In steps 525 to 536, the number of evaluations NE and the first determination value NEd
1, second judgment value NEd2 success index RSC, success rate judgment value RSCd, RSC
e, RSCf is used to correct the selection reference values DSTb1 and DSTb2. Here, for example, NEd1 = 500, NEd2 = 800, RSCd =
0.10, RSCe = 0.05, RSCf = 0.05. In step 283 of the initial setting program 28 (FIG. 13), for example, SELS
= 1, DSTb1 = 250, DSTb2 = 250.

Now, in step 525, when it is determined that NE ≦ NEd1, that is, the search is still in the “first half”, in step 527,
ELS is set to 1, and the first selection condition is specified as the cross pair selection condition. Then, in step 529, the success index RSC is compared with the success rate determination value RSCd. When the success index RSC is lower than the success rate determination value RSCd (RSC ≦ RSCd), the value of the currently set first selection criterion value DSTb1 is too strict, and the number of sets meeting the conditions is small and the success index RSC is It is determined that it does not become high, and the first selection reference value DSTb1
Is set slightly smaller (for example, 5%).

On the other hand, when NE ≧ NEd2, that is, when the search is “end stage”, the process proceeds to steps 525 → 526 → 528, SELS is set to 2, and the second selection condition is designated as the cross pair selection condition.
Then, in step 529, the success index RSC and the success rate judgment value RS
Compare with Cf. If the success index RSC is lower than the success rate determination value RSCf (RSC ≦ RSCf), the currently set second
It is determined that the number of crossover sets meeting the conditions is small and the success indicator RSC does not increase because the value of the selection criterion value DSTb2 is too severe. In step 532, the value of the second selection criterion value DSTb2 is slightly reduced (for example, 5% ).

When the success index RSC is higher than the success rate determination value RSCf, it is determined that the selection criterion value of the currently selected cross pair selection condition is not a problem, and the processing of the selection condition correction program 52 is directly performed. finish.

If it is determined in steps 525 and 526 that NEd1 <NE <NEd2, that is, “middle”, the success index RSC is compared with a success determination value RSCe in step 533. Success indicator R
When SC is lower than the success determination value RSCe (RSC ≦ RSCe),
Since the current selection condition is considered inappropriate, it is necessary to switch the selection condition. Therefore, step 534
After it is determined whether the current selection condition is the first or second, the selection condition is changed to the second selection condition in step 535 (SELS =
2) Or, in step 536, the selection condition is changed to the first selection condition (SELS = 1).

When the success index RSC is higher than the success determination value RSCe, it is determined that the currently selected selection condition is not a problem, and the process of the selection condition modification program 52 is terminated.

As described above, in the eleventh embodiment, the inter-set distance indicating the similarity between the two sets is used as the selection criterion of the crossed pair. Therefore, the characteristics of both the global search and the local search are used. To generate a new set.

In other words, by preferentially selecting a pair in which the distance between sets is equal to or greater than the first selection reference value DSTb1 and performing crossover between two sets having different characteristics as much as possible, there are many hits and misses and the convergence deteriorates. However, the probability that a set having better group management performance is generated can be increased. on the other hand,
If a pair whose distance between sets is equal to or less than the second selection reference value DSTb2 is preferentially selected and crossover is performed between two sets having similar characteristics as much as possible, a set having extremely excellent group management performance can be obtained. Is less likely to be generated, but the probability of generating a new set having reasonably good group management performance can be increased.

In the eleventh embodiment, particularly, the initial or final stage of the search is determined based on the number of evaluations, and in the initial stage of the search, a pair in which the distance between sets is equal to or larger than the first selection reference value DSTb1 is preferentially selected to achieve a wide area. At the end of the search, on the other hand, at the end of the search, a pair in which the inter-set distance is equal to or less than the second selection reference value DSTb2 is preferentially selected to emphasize the convergence of the search, so that the search efficiency can be improved.

In the eleventh embodiment, particularly, the distance between sets is the first.
In the case where the search is performed using the cross pair selection condition of preferentially selecting a pair having the selection criterion value DSTb1 or more, if the success index RSC becomes lower than the expected value at the beginning of the search, the crossover is performed. The first selection criterion value DSTb1 is set to be smaller than the current value so that the pair selection condition becomes milder than the current value.
When the search is sluggish due to inappropriate DSTb1, the appropriate first selection reference value DSTb1 can be automatically set to overcome the sluggish state.

In the eleventh embodiment, particularly, the distance between sets is the second
When a search is performed using the crossed pair selection condition of preferentially selecting a pair having a selection criterion value DSTb2 or less, at the end of the search, the value of the success indicator RSC is higher than the expected value. Is lower, the second selection criterion value DSTb2 is set to be larger than the current value so that the cross pair selection condition becomes milder than the current value. Therefore, the set second selection criterion value DSTb2 is inappropriate. When the search is sluggish, the appropriate second selection reference value DSTb2 is automatically set, and the sluggish state can be overcome.

In the eleventh embodiment, especially, in the middle stage of the search, the first
If you are searching using cross-pair selection conditions,
If the value of the success indicator RSC is lower than expected,
Since the selection condition is changed to the second cross pair selection condition, if the selection condition does not adapt to the current situation and the search is sluggish,
It is possible to automatically switch to an appropriate selection condition.

Further, in the eleventh embodiment, particularly, at the end of the search, the second
If you are searching using cross-pair selection conditions,
If the value of the success indicator RSC is lower than expected,
Since the selection condition is changed to the first cross pair selection condition, if the selection condition does not adapt to the current situation and the search is sluggish,
It is possible to automatically switch to appropriate selection conditions.

[Embodiment 12 (Other Embodiment of Crossover Pair Selection)] Another embodiment of the selection condition correcting unit 5 will be described with reference to FIG. FIG. 40 shows a selection condition correction program.
This is a modification of a part of the selection condition modification program 52 (see FIG. 39) of the eleventh embodiment.

In FIG. 40, in step 522, the number of evaluations is a fixed number N
It is determined that it is Eb (for example, 50 times). Step 52
In step 6, it is determined whether or not it is the second period (end stage). If the period is other than the second period (NE <NEd2), the first condition is specified as the cross pair selection condition in step 527, and the first selection criterion is determined in step 531. The value DSTb1 is set slightly smaller (for example, 2%). On the other hand, in step 526, the second period (ie, NE
If it is determined that ≧ NEd2), in step 528, the second condition is designated as the crossover set selection condition, and in step 532
To set the second selection reference value DSTb2 a little smaller (for example, 2%).

As described above, in the twelfth embodiment, the first selection reference value DSTb1
Can be switched in accordance with the search progress within the period in which the cross pair selection condition using is used. That is, since the value of the first selection reference value DSTb1 at the beginning of the period is set to be larger than the value at the end of the period, the diversity of the group management performance is emphasized at the beginning of the period, and at the end of the period, Can emphasize search convergence.

Similarly, in the twelfth embodiment, the second selection reference value DSTb2
Can be switched in accordance with the search progress within the period in which the cross pair selection condition using is used. That is, since the value of the second selection reference value DSTb2 at the beginning of the period is set to be smaller than the value at the end of the period, the diversity of group management performance is emphasized at the beginning of the period, and at the end of the period, Can emphasize search convergence.

Therefore, the search efficiency can be further improved by switching the selection conditions finely in this way.

Embodiment 13 (Another Embodiment of Selection of Crossover Parameters) In the first embodiment, crossover parameters (parameter positions) at which parameter values are exchanged are randomly selected for two parent sets. The generation unit of the thirteenth embodiment is characterized in that a parameter value difference (parameter deviation) is used as a parameter selection condition, and further, the parameter selection condition is modified according to a search situation.

The generation unit will be described with reference to FIGS. 41 and 42.
The thirteenth embodiment will be described focusing on the differences from the eleventh embodiment.

FIG. 41 shows a new set generation program 31 (see FIG.
14) is shown.
In FIG. 41, in step 318a, the value of the counter RC is 0
Initially set to In this embodiment, the counter RC is used to count the number of times the parameter deviation condition, which is one of the cross parameter selection conditions, is determined. In step 318b, one random number is generated from 0 to 25, and the parameter number PX is specified by the value of the random number. This is the same as in the first embodiment.

In the next step 318c, the number of times the above parameter deviation condition is determined is compared with a predetermined number. Step 318d-31
If the crossover parameter PX that satisfies the parameter deviation condition cannot be found even after repeating 8h a predetermined number of times (for example, 10 times) or more, it is determined that the selection in consideration of the parameter deviation condition has been sufficiently performed, and the process of step 318 is performed. The processing ends, and the parameter number PX selected in step 318b is determined as the number of the crossover parameter.

On the other hand, in step 318c, if the number of times the parameter deviation condition is determined is less than the predetermined number (RC <10), step 31
The process proceeds from 8c to 318d, where the counter RC is incremented by one.
In the next step 318e, the two selected sets PS1, PS
The difference between the PX-th numerical values in 2 [| EPS (PS1) <PX
> −EPS (PS2) <PX> |], and this is set as the distance DSTP.

It is assumed that each parameter value is converted into a numerical value between 0 and 100 and normalized. Step 387 of the optimal set extraction program 38 (see FIG. 20)
When the optimum set data BPD is created, the value is returned to a value that can be used by the group management device 1.

As described above, in step 318e, the two sets PS1, PS
When the difference DSTP of the second PX-th parameter value is calculated, it is next determined in steps 318f to 318h whether or not the deviation condition selected and designated is satisfied.

The first selection condition is specified as the parameter deviation condition (SELS
= 1), the process proceeds to step 318f → 318g, where it is determined whether the deviation DSTP is equal to or greater than the first selection reference value DSTc1. When DSTP <DSTc1, the crossover parameter PX that does not satisfy the parameter deviation condition is discarded, the process returns to step 318b, and the same process is repeated from the beginning.

The second selection condition is specified as the parameter deviation condition (SELS
= 2), the deviation is similarly calculated at step 318h.
It is determined whether DSTP is equal to or less than a second selection criterion value DSTc2. If the selection condition is not satisfied (DSTP> DSTc2), step 3 is performed.
I will go back to 18b and start over.

Steps 318b to 318b are performed until the parameter deviation condition is determined a predetermined number of times (10 times) or more, or until a cross parameter that satisfies the parameter deviation condition is found.
The operation of 8h is repeated. If a crossover parameter PX that satisfies DSTP ≧ DSTc1 or DSTP ≦ DSTc2 is detected on the way, it is determined as a normal crossover parameter PX, and the process proceeds to the next step 319. The procedure after step 319 is Example 1
Therefore, the description is omitted.

A method of correcting the above selection condition according to the search situation will be described with reference to FIG.

FIG. 42 shows Step 5 of the calculation program 100 (see FIG. 37).
FIG. 9 is a diagram showing a selection condition modification program of FIG.

The selection condition correction program in the thirteenth embodiment is the same as the selection condition correction program 52 (FIG. 39) in the eleventh embodiment, except for steps 531 and 532. Therefore, description will be made focusing on the operations of steps 529 to 532 for correcting the selection reference values DSTc1 and DSTc2. Note that the initial setting program 28 (Fig. 13
It is assumed that, in step 283), the designation data SELS = 1, the selection reference value DSTc1 = 50, and the selection reference value DSTc2 = 50.

At the beginning of the search (NE ≦ NEd1), if the first selection condition is being selected (SELE = 1) and the success indicator RSC is lower than the expected value RSCd (RSC ≦ RSCd), the currently set first selection condition is set. Because the value of the selection reference value DSTc1 is too strict and inappropriate,
It is determined that the number of crossover parameters meeting the condition is small and the success index RSC does not become high, and the process proceeds to step 529 → 531, where the value of the first selection criterion value DSTc1 is set slightly smaller (for example, 5%). The processing of the selection condition modification program 52 ends.

On the other hand, at the end of the search (NE ≧ NEd2), if the second selection condition is currently selected (SELS = 2) and the success index RSC is lower than the expected value RSCf (RSC ≦ RSCf), the current setting is made. Since the value of the second selection criterion value DSTc2 is too severe and inappropriate, it is determined that the number of crossover parameters meeting the conditions is small and the success index RSC does not become high, and steps 530 → 532
Then, the value of the second selection reference value DSTc2 is set slightly larger (for example, 5%).

If the success index RSC is higher than the expected value RSCd, it is determined that the selection reference value of the currently selected selection condition does not have any problem with the current value, and the processing of the selection condition correction program 52 ends as it is. .

As described above, in the thirteenth embodiment, the parameter deviation indicating the similarity between the two selected sets is obtained, and the crossover parameter selection condition is set based on the parameter deviation. Generation can be performed in consideration of both the convergence of search.

That is, the parameter deviation is equal to the first selection reference value DSTc.
If a control parameter that is 1 or more is selected preferentially and parameters whose characteristics are as far apart as possible are crossed over, a new set with very good group management performance, although the convergence of search will be worse due to many hits and misses Is likely to be generated. In addition, the parameter deviation is the second
By preferentially selecting control parameters that are equal to or less than the selection criterion value DSTc2, and crossing parameters having similar characteristics as much as possible, the possibility of generating a new set having extremely excellent group management performance is reduced, The probability that a new set having reasonably good group management performance is generated can be increased.

In the thirteenth embodiment, the initial and final stages of the search are determined based on the number of evaluations. At the beginning of the search, a control parameter having a parameter deviation equal to or more than the first selection reference value DSTc1 is preferentially selected to improve the group management performance. A search emphasizing diversity can be performed. On the other hand, at the end of the search, a control parameter whose parameter deviation is equal to or less than the second selection reference value DSTc2 is preferentially selected to perform a search emphasizing the convergence of the search. be able to. Therefore, it is possible to generate a new set in consideration of both the diversity of the group management performance and the convergence of the search according to the search time.

In the thirteenth embodiment, when the search is performed using the cross parameter selection condition using the first selection criterion value DSTc1, the value of the success indicator RSC is lower than the expected value at the beginning of the search. When the first selection criterion value DSTc1 is lower than the current value, the first selection criterion value DSTc1 is set to be smaller than the current value so that the selection condition becomes milder than the current selection value.
If the search is sluggish due to inappropriate setting of Tc1,
The value can be automatically corrected to an appropriate value.

Further, in the above-described embodiment 13, when the search is performed using the cross parameter selection condition using the second selection criterion value DSTc2, at the end of the search, the value of the success indicator RSC is smaller than the expected value. Is also set lower than the current value, the second selection criterion value DSTc2 is set to be larger than the current value so that the selection condition becomes milder than the current selection value.
If the search is sluggish due to inappropriate setting of Tc2,
The value can be automatically corrected to an appropriate value.

In the above-described embodiment 13, in the middle stage of the search, the first
When the search is performed using the first crossover parameter selection condition using the selection criterion value DSTc1, the success index RSC
Is changed to the second crossover parameter selection condition using the second selection criterion value DSTc2, the first crossover parameter selection condition is inappropriate for the current situation. If the search is sluggish, it is possible to automatically switch to an appropriate selection condition.

In the above-described embodiment 13, in the middle stage of the search, the second
When the search is performed using the second crossover parameter selection condition using the selection criterion value DSTc2, the success indicator RSC
Is changed to the first crossover parameter selection condition using the first selection criterion value DSTc1, the second crossover parameter selection condition is inappropriate for the current situation. If the search is sluggish, it is possible to automatically switch to an appropriate selection condition.

As described above, by changing the selection criterion value or switching the selection condition according to the search situation, it is possible to appropriately perform a break from the search sluggish state and improve the search efficiency.

Embodiment 14 (Another Embodiment of Selection Condition Correction) Next, another embodiment of the selection condition correction unit 5 will be described with reference to FIG. FIG. 43 is a diagram showing an operation procedure of the selection condition correction program 52, and the selection condition correction program 52 of the thirteenth embodiment.
(See FIG. 42).

In FIG. 43, in step 522, it is determined that the number of times of evaluation has reached a predetermined value NEb (for example, 50 times). In step 526, it is determined whether or not it is the second period (end stage),
If it is not the second period (NE <NEd2), the first condition is specified as the crossover parameter selection condition in step 527, and the first selection reference value DSTc1 is slightly reduced (for example, 2%) in step 531.
Set only smaller.

On the other hand, in step 526, the second period (that is, NE ≧ NEd)
If it is determined to be 2), at step 528, the second condition is designated as the crossover parameter selection condition, and at step 532
Then, the second selection reference value DSTc2 is set slightly smaller (for example, 2%).

As described above, the condition regarding the parameter deviation is changed according to the search situation. Operations other than the above are described in the embodiment.
Exactly the same as 13.

As described above, in the fourteenth embodiment, the value of the first selection criterion value DSTc1 at the beginning of the period is changed to the value at the end of the period when the cross parameter selection condition using the first selection criterion value DSTc1 is used. Larger than
Since the above selection conditions are set so that the beginning of the period becomes stricter than the end, the diversity of the group management performance can be emphasized at the beginning of the period, and the convergence of the search can be emphasized at the end.

In the above-described embodiment 14, during the period in which the cross parameter selection condition using the second selection reference value DSTc2 is used, the value of the second selection reference value DSTc2 at the end of the period is set to be smaller than the initial value. However, since the above selection conditions are set so that the end of the period becomes stricter than the beginning, the diversity of the group management performance can be emphasized at the beginning of the period, and the convergence of the search can be emphasized at the end of the period. .

In Examples 9 to 14, the first half / second half of the search, the initial / final stage, and the like were determined in accordance with the number of evaluations NE. Instead of the number of evaluations NE, the number of additional registrations NR may be used to determine the first half / second half of the search, the initial / final stage, and the like.

Fifteenth Embodiment (Another Embodiment of Parameter Selection) Next, another embodiment of the generation unit will be described with reference to FIGS. 44 and 45. In the fifteenth embodiment, the selection probability (appearance rate) for each parameter is based on both the degree of association with the traffic flow characteristics and the degree of association with the evaluation item of the group management performance.
Is set. The basic configuration of the fifteenth embodiment is the same as that of the second embodiment, and therefore, the description will focus on the differences from the second embodiment.

FIG. 44 shows a new set generation program 31 (see FIG. 14).
Of step 318 in FIG.

In FIG. 44, in steps 318j to 318q, the parameter appearance rate RP for 25 parameters according to the traffic flow specification
Set A (1) to RPA (25). Specifically, first, in steps 318j to 318l, the type of the traffic flow is determined based on the contents of the total number of passengers, the entrance floor traffic ratio, the up traffic ratio, the down traffic ratio, and the like included in the traffic flow specification data TRS. to decide. In other words, if the current traffic conditions are
It is determined which state is in the down peak and the normal time zone.

If it is determined that the time period is the normal time zone, in step 318m, the appearance rates RPA1 (1) to RPA1 (25) for each parameter prepared for the normal time zone in advance are set to the parameter appearance rate RPA.
(1) Set as RPA (25). Similarly, during the work hours, in steps 318n, RPA2 (1) to RPA2
(25) is set as the parameter appearance rate RPA (1) to RPA (25), and when it is up-peak, RPA3 is set in step 318p.
(1)-RPA3 (25) is the parameter appearance rate RPA (1)-RPA
Set as (25) and step 31 for down peak
Parameter appearance rate RPA4 (1) to RPA4 (25) in 8p
(1) Set as RPA (25).

The parameter appearance rates RPA1 (1) to RPA1 (25), RPA2 (1) to RPA2 (25) prepared for each traffic flow are shown in 10B of FIG.
RPA3 (1) to RPA3 (25) and RPA4 (1) to RPA4 (2
5) is shown.

In FIG. 45, when the characteristic of the traffic flow is a normal time zone, the parameters (parameter numbers = 1 to 9, 22) having a high relation with the traffic flow are related to the parameter appearance rates RPA1 (1) to RPA1 (25). 25 to 25), the appearance rate is set to 10, and the appearance rate is set to 0 for almost irrelevant parameters (parameter numbers = 18 to 21). The appearance rate is set to 5 for a control parameter having a moderate degree of association (parameter number = 10 to 17).

When the characteristic of the traffic flow is the work time zone, regarding the parameter appearance rates RPA2 (1) to RPA2 (25), a control parameter having a high relevance to the traffic flow (parameter number = 1)
~ 9, 18 ~ 21), the appearance rate is set to 10, and the control parameters (parameter number = 10 ~
For 17), the appearance rate is set to 0. In addition, a control parameter having a moderate degree of relevance (parameter number =
22 to 25), the appearance rate is set to 5.

When the characteristics of the traffic flow are up-peak, the control parameters (parameter number = high) related to the traffic flow are related to the parameter appearance rates RPA3 (1) to RPA3 (25).
The appearance rate is set to 10 for 1 to 13), and the appearance rate is set to 0 for almost unrelated control parameters (parameter numbers = 14 to 21). In addition, control parameters having a moderate degree of relevance (parameter number = 22 to 2
For 5), the appearance rate is set to 5.

When the characteristics of the traffic flow are down-peak, the control parameters (parameter number = 1) that are highly relevant to the traffic flow with respect to the parameter appearance rates RPA4 (1) to RPA4 (25)
~ 9, 14 ~ 17), the appearance rate is set to 10, and the control parameters (parameter number = 18 ~
For 21), the appearance rate is set to 0. In addition, a control parameter having a moderate degree of relevance (parameter number =
22 to 25), the appearance rate is set to 5.

As described above, in steps 318j to 318q, the parameter appearance rates RPA (1) to RPA (25) for 25 control parameters are set according to the traffic flow specifications. In addition, the appearance rates RPA1 (1) to RPA1 (25), RPA2 (1) to RPA2 (2
5), RPA3 (1) to RPA3 (25), and RPA4 (1) to RPA4
The value of (25) is not limited to the value shown in FIG. Any value may be set as long as it relatively represents the degree of association with each traffic flow characteristic. Further, the appearance rate may be more finely differentiated between the control parameters.

Returning to FIG. 44, next, in step 318r, the parameter appearance rate is calculated using correction values RPAA (1) to RPAA (25) proportional to the degree of association with the evaluation item (for example, average waiting time) of the group management performance. RPA (1) to RPA (25) are corrected.

Here, the correction values RPAA (1) to RPAA (25) are set by the performance reference value setting device 3 described above. Here, as described in the second embodiment, the performance reference value setting device 3 outputs the “target value” for the average waiting time and the “designated value” for the evaluation reference value BX, as in the second embodiment. In the second embodiment, in addition, the “degree of association” with the average waiting time, which is an evaluation item, is output as a correction value.

Therefore, the reference value data TGT in the search condition signal 1a input from the group management device 1 to the search device 10 includes the waiting time target value TAW, the additional reference designation value TCB, and the correction value RPAA (1) to
RPAA (25) is included. RPAA (1) to RPAA (25) in FIG.
Indicates a correction value for the target value of the average waiting time.

In the correction values RPAA (1) to RPAA (25) of FIG. 45, the correction value is 10 for the control parameter (parameter number = 8, 22, 23) which is highly relevant to the average waiting time as the evaluation item.
, And the correction value is set to 0 for a control parameter (parameter number = 10 to 21) having little relation. Further, the appearance rate is set to 5 for control parameters having moderate relevance (parameter numbers = 1 to 7, 9, 24, and 25).

On the other hand, if the evaluation item is related to power saving instead of the average waiting time, the correction values RPAA (1) to RPAA (25)
Relevant control parameters (parameter number = 4 to
7, 22 to 25), the correction value is set to 10, and the control parameter (parameter number = 9 to
For 21), the correction value is set to 0. In addition, control parameters having a moderate degree of relevance (parameter number = 1 to
For 3 and 8), the correction value is set to 5.

Similarly, in the case of evaluation items other than the above, correction values RPAA (1) to RPAA (25) are set according to the degree of association. Note that the correction values RPAA (1) to RPAA (25) may be set to any values as long as they relatively represent the degree of association with the evaluation item. In addition, the correction value may be set with a finer difference between the control parameters.

Next, in step 318s, one random number having a value between [0] and [the total value of the parameter appearance rates RPA (1) to RPA (25)] is generated, and crossover or Determine the number PX of the parameter at which the mutation is made. Then, the process proceeds to the next Step 319. The procedure after step 319 is the same as in the first embodiment, and a description thereof will not be repeated.

As described above, in the fifteenth embodiment, since the degree of association between the parameter and the traffic flow characteristic is used as the parameter selection condition, the value of the parameter closely related to the specific traffic flow characteristic can be preferentially changed. , The possibility that a new set having excellent group management performance is generated can be increased.

Further, in Embodiment 15, the appearance rate is set in proportion to the degree of association with the traffic flow characteristic for each parameter, and the parameter is selected according to this appearance rate. The parameters that easily affect the management performance are more easily selected, and the possibility of generating a new set having better group management performance can be increased.

In the above-described embodiment 15, the appearance rate is set to 0 for control parameters that are not related to the traffic flow characteristics when performing group management.
Is set so as not to be selected, so that the application of crossover or mutation to irrelevant parameters can be completely prevented.

In Embodiment 15, the degree of association between the parameter and the evaluation item to be evaluated is used as the parameter selection condition, so that the value of a parameter closely related to the evaluation item is preferentially changed, and thus, excellent performance is achieved. The possibility of generating a new set having group management performance can be increased.

Further, in Embodiment 15 described above, for each parameter, an appearance rate proportional to the degree of association with the evaluation item to be evaluated is set, and the parameter is selected in accordance with this appearance rate. The parameters that are more likely to influence are more likely to be selected, and can increase the possibility of generating a new set having better group management performance.

In the fifteenth embodiment, the appearance rate is set to 0 for the control parameter not related to the evaluation item to be evaluated so that the control parameter is not selected. Can be completely prevented.

Further, in the above-described Embodiment 15, the parameter condition is a combination of the degree of association between the parameter and the evaluation item to be evaluated, and the degree of association between the control parameter and the traffic flow characteristic. It is possible to increase the possibility of generating a new set.

As described above, according to the fifteenth embodiment, unnecessary new set generation, evaluation, additional registration determination, and the like can be reduced, and search can be performed efficiently.

When a plurality of evaluation items are included in the evaluation function of the group management performance, a weight or a correction value of the appearance rate may be set according to the importance of each evaluation item.

[Embodiment 16 (Other Embodiment of Crossover Pair Selection)] Next, another embodiment of the generation unit will be described. In this embodiment, an intersection set is selected based on the number of similar sets. Hereinafter, the sixteenth embodiment will be described focusing on parts different from the first embodiment.

In the new set generation program 31 (see FIG. 11),
In the sixteenth embodiment, the process of selecting the crossover pairs PS1 and PS2 in step 317 (see FIG. 14) is significantly different from that of the first embodiment, and this will be described with reference to FIG.

In FIG. 46, first in step 317j, the above [3
1] distance DST (i, j) (where i, j
= 1, 2,..., P, i ≠ j). This calculation is performed in step 414 of the deletion program 35 (FIG. 26) of the third embodiment.
Is the same operation as. It is assumed that the value of each parameter is normalized to a value between 0 and 100.

Then, in step 317k, the set number i is initialized to 1, and in step 317p, the setting of the appearance rates RSA (1) to RSA (P) is completed for all the sets (i = 1, 2,..., P). Until this is detected, the processing of steps 317l to 317n is repeated. In step 317l, all j = 1, 2,..., P
, The number of sets for which DST (i, j) ≤ DSTa
(I) Obtain (the number of similar sets). Here, DSTa is 2
This is a determination value for determining whether two sets are similar to each other.
Is set.

Next, in step 317m, based on the number of similar sets MDST (i), the appearance rate RSA (i) of the set i is calculated as RSA (i) =
The calculation is performed using the formula 1 {MDST (i) +1}. That is,
The smaller the number of similar sets, the higher the appearance rate. Then, in order to calculate the appearance rate for the next set, the set number i is increased by 1 in step 317n.

Thus, the appearance rates of all sets RSA (1) to R
After setting SA (P), finally, in step 317q, [0]
, Generate two random numbers between [the sum of the appearance rates RSA (1) to RSA (P)], and calculate the values of each random number and the appearance rate RUA.
(1) Two parent sets PS1 and PS2 according to RUA (P)
Select

Thus, the processing of step 317 is completed, and the two sets PS1 and PS2 are determined as they are as normal cross pairs.
Proceed to the next step 318. The procedure after step 318 is the same as in the first embodiment, and a description thereof will be omitted.

As described above, in the sixteenth embodiment, since the crossing pairs are selected based on the number of similar sets, the probability of crossing pairs having different characteristics from each other can be improved. Therefore, the probability of generating a new set having very excellent group management performance can be improved.

In each of the above embodiments, only one crossover parameter is selected (a method for selecting such a parameter is generally called one point crossover). However, the present invention is not limited to this. . A method of selecting two or more crossover parameters simultaneously (multi-point crossover)
over)) can also be adopted. In addition, a bit string (mask) having the same length as the number of parameters is prepared in advance, and which parent gene (parameter value) is passed to the child is determined based on the value of each bit specified by the mask. A method called uniform crossover may be adopted. This is the same for “mutation”.

Although the above set is composed of 25 parameters, the number and contents are merely examples, and the present invention can be applied to various parameter value sets used in the group management algorithm.

Embodiment 17 (Other System Configuration Examples) In each of the above embodiments, the group management device 1 and the search device 10 are provided in the elevator machine room of the building, and the optimum set is determined online. However, it is not limited to this.

For example, regarding the above Examples 1 to 6 and 8 to 15,
As shown in FIG. 47, the search device 10 and the simulation device 2 are installed in a monitoring center of an elevator maintenance company, and the communication device 4A and the communication device communicate between the search device 10 and the group management device 1.
It is also possible to connect with a telephone line using 4B. In this case, the communication device 4A can perform data communication with another building having the communication device 4B. With this configuration, one set of search device and simulation device can be shared by a plurality of group management devices. In addition, the search device 10, the simulation device 2, and the communication device 4A can be installed in the manager's office or the disaster prevention center of the building.

According to the seventeenth embodiment, the expensive search device 10 and the simulation device 2 are shared, and the system can be configured at low cost.

By the way, as for the eighth embodiment, as shown in FIG. 48, the search device 10 is installed in a management room of a building or a monitoring center of an elevator maintenance company, and the search device 10 and the group management device 1 are installed.
It is also possible to establish a connection between them by a telephone line using the communication device 4A and the communication device 4B.

The search device 10 can also be used for developing a group management algorithm, that is, when selecting an optimum group management algorithm plan from a plurality of group management algorithm plans. Normally, when developing a new group management algorithm, a simulation is performed using a simulation device, and based on the group management performance data PRF obtained at that time, the performance of the group management algorithm is evaluated or an optimal set is determined. Is done. In this case, the search device 10 is connected to the simulation device 2 as shown in FIG.

Further, when the group management device 1 is shipped from the factory, the search device 10 searches for and registers the optimum set by the simulation device 2 in FIG.
It can be used when registering S1 to GPS4.

The search device 10 and the simulation device 2 can be realized by one microcomputer instead of being configured separately, and the group management device 1, the search device 10, and the simulation device 2 are configured by one microcomputer You can also.

──────────────────────────────────────────────────続 き Continuation of the front page (56) References JP-A-7-291541 (JP, A) JP-A-6-305649 (JP, A) Labor and two others, “Elevator group management system using genetic algorithm Constitution Method ", Proc. Of the Conference on Systems, Control and Information Engineers, 1992, vol. 36th, p. 85-86 (58) Field surveyed (Int. Cl. ⁷ , DB name) B66B 1/00-1/52

Claims (37)

(57) [Claims] 1. A system for group-managing a plurality of elevators according to a group-management algorithm including a plurality of parameters, wherein a search is performed to find an optimum set from a set that is a combination of parameter values given to the group-management algorithm. Storage means for storing a plurality of sets, one or more sets selected as a parent from the storage means, and one or more new sets partially inheriting the properties of the parent. Generating means for generating a set; evaluating means for obtaining an execution result when the group management algorithm is executed using the new set as a group management performance value; adding the new set to the storage means; And deleting the inferior set from the storage means, together with a selection means for improving a plurality of sets stored in the storage means, Extracting means for extracting the optimum set from the plurality of sets stored and improved in the storage means based on the group management performance value; and prerequisites for a search given at the start of the search for the optimum set An elevator group management system, comprising: re-search determination means for instructing the search device to re-search for the optimal set based on the change of. 2. The system according to claim 1, wherein the generation unit generates two new sets by exchanging some numerical values between the two sets selected from the storage unit. New value replacing means for generating one new set by replacing some parameter values in one set selected from the storage means with new random numbers generated. An elevator group management system, comprising: a generation method selection unit that selects the exchange and the replacement stochastically. 3. The system according to claim 1, wherein said generating means includes: a parent selecting means for selecting one or two sets from said storage means; and exchanging numerical values or replacing new values for said one or two sets. Parameter selection means for selecting parameters to be performed, and two new sets by exchanging the values of the parameters selected by the parameter selection means between the two sets selected by the parent selection means. Generating a new set by replacing the value of the parameter selected by the parameter selecting means in the one set selected by the parent selecting means with a new numerical value generated at random; New value replacement means for generating a replacement method; and a generation method selection means for stochastically selecting the exchange and the replacement. Elevator group management system comprising: the means. 4. The system according to claim 3, wherein said parent selecting means selects a parent based on parent selection criterion information for increasing a probability that an excellent new set is generated. Elevator group management system. 5. The system according to claim 4, wherein said parent selection criterion information is an inter-set distance, said parent selection means calculates said inter-set distance, and said set distance is determined from said storage means. An elevator group management system, wherein a pair of sets satisfying a condition is randomly selected. 6. The system according to claim 4, wherein said parent selection criterion information is said group management performance value, and said parent selection means weights a selection probability to each set according to said group management performance value. An elevator group management system, wherein one or two sets are randomly selected from the storage means. 7. The system according to claim 4, wherein the parent selection criterion information is the number of similar sets, and the parent selecting means calculates the number of similar sets for each set, and based on the number of similar sets. An elevator group management system, wherein each set is weighted with a selection probability, and one or two sets are randomly selected from the storage means. 8. An elevator group management system according to claim 3, further comprising means for correcting a parent selection condition according to a progress of the search. 9. The system according to claim 3, wherein said parameter selecting means selects a parameter based on parameter selection criterion information for increasing a probability that an excellent new set is generated. Elevator group management system. 10. The system according to claim 9, wherein the parameter selection criterion information is a difference between two parameter values exchanged between the two sets, and wherein the parameter selection means calculates the difference. An elevator group management system, wherein a parameter whose difference satisfies a predetermined condition is randomly selected. 11. The system according to claim 9, wherein the parameter selection criterion information is a degree of association between an elevator use situation and each parameter, and the parameter selecting means includes:
An elevator group management system, wherein each parameter is weighted with a selection probability, and the parameter is selected at random. 12. The system according to claim 9, wherein the parameter selection criterion information is a degree of association between the content of the performance evaluation value and each parameter, and the parameter selecting means, in accordance with the degree of association,
An elevator group management system, wherein each parameter is weighted with a selection probability, and the parameter is selected at random. 13. The elevator group management system according to claim 3, further comprising means for correcting a parameter selection condition according to a progress of the search. 14. The elevator group management system according to claim 2, further comprising probability correction means for correcting the selection probability of each generation method according to the progress of the search. 15. The system according to claim 14, wherein said probability correcting means calculates a success index from a ratio between the number of sets evaluated and the number of sets added to said storage means, and based on the success index. An elevator group management system, wherein the selection probability is corrected. 16. A system for managing a plurality of elevators in accordance with a group management algorithm including a plurality of parameters, wherein the search device searches for an optimum set from among sets which are combinations of parameter values given to the group management algorithm. The search device includes: a storage unit for storing a plurality of sets; and two parameter sets selected as parents from the storage unit. Numerical exchange means for generating two new sets partially inheriting the properties of the parent, and new numerical values randomly generated from a part of parameter values in one set selected as a parent from the storage means New value replacing means for generating one new set partially inheriting the property of its parent by replacing Generating method selecting means for stochastically selecting the exchange and the new value replacement; evaluating means for obtaining an execution result when the group management algorithm is executed using the new set as a group management performance value; An adding means for additionally storing only a good new set satisfying the addition condition in the storage means; a deletion means for deleting from the storage means a bad set satisfying a predetermined deletion condition; and Extracting means for extracting the optimum set from the plurality of sets based on the group management performance value; and performing a search based on a change in a precondition of a search given at the start of the search for the optimum set. An elevator group management system, comprising: re-search determination means for instructing a device to re-search for the optimum set. 17. The elevator group management system according to claim 16, further comprising means for correcting said additional condition. 18. The elevator group management system according to claim 17, wherein said additional condition is determined based on a group management performance value of each set, and said additional condition is gradually set strictly. 19. The elevator group management system according to claim 16, wherein said deletion means deletes a set whose performance evaluation value is inferior. 20. The elevator group management system according to claim 19, wherein said deleting means deletes a set similar to another set based on a distance between sets. 21. The elevator group management system according to claim 16, further comprising initial setting means for performing initial setting of a search. 22. The system according to claim 21, wherein said initial setting means includes a first initial setting mode and a second initial setting mode. In said first mode, initializing is performed using a plurality of sets prepared in advance. In the second mode, initial setting is performed using a plurality of sets improved in the previous search, and the first and second initial setting modes and the second initial setting mode are set according to the search start condition. The elevator group management system, wherein is selected. 23. The elevator group management system according to claim 16, further comprising end determination means for determining a search end according to a search situation. 24. The elevator group management system according to claim 23, wherein said end determination means determines the end of the search based on the number of evaluated sets. 25. The elevator group management system according to claim 23, wherein said end determination means determines the end of the search based on the number of added sets. 26. The system according to claim 23, wherein said end judging means judges the end of the search based on a success index which is a ratio between the number of evaluated sets and the number of added sets. Elevator group management system. 27. The system according to claim 23, wherein the end determining means calculates an inter-set distance for a plurality of sets stored in the storage means, and determines the end of the search based on the inter-set distance. An elevator group management system, characterized in that: 28. The system according to claim 16, wherein the prerequisites for the search include elevator specification data representing a plurality of elevator specifications and traffic flow specification data on passenger traffic. Group management system. 29. The elevator group management system according to claim 16, wherein said storage means further stores said group management performance value associated with each set. 30. The elevator group management system according to claim 16, wherein a target value setting device is connected to the search device for setting a target value related to the search. 31. The system according to claim 16, wherein the search device includes the group management algorithm, the group management device controls operation of a plurality of elevators, and a group management algorithm included in the group management device. And a simulation device including the same group management algorithm, wherein the evaluation unit sets an execution result of the simulation device as a group management performance value. 32. The system according to claim 31, wherein the search device and the simulation device are remotely located with respect to the group management device, and the search device and the group management device are interconnected by a communication line. An elevator group management system characterized by the following. 33. The system according to claim 16, wherein the search device is connected to a group management device that includes the group management algorithm and controls the operation of a plurality of elevators. An elevator group management system, wherein an execution result when a group management algorithm is simulated is used as the group management performance value. 34. The system according to claim 33, wherein the search device is remotely located with respect to the group management device,
An elevator group management system, characterized in that these are interconnected by a communication line. 35. A system for group-managing a plurality of elevators according to a group management algorithm including a plurality of parameters, wherein the system controls operation of a plurality of elevators installed in a building according to the group management algorithm. , Elevator specification data representing the specifications of the plurality of elevators,
A group management device that outputs traffic flow specification data about passenger traffic in the building; andthe elevator management data and the traffic flow specification data output from the group management device; A search device that searches for an optimal set as a set that is a combination of parameter values to be given to the group management algorithm, and a simulation that is connected to the search device and includes the same group management algorithm as the group management algorithm of the group management device. And a storage device for storing a plurality of sets, and exchanging some parameter values between the two sets selected as parents from the storage device. Means for generating two new sets partially inheriting the properties of the parent, By replacing some parameter values in one set selected as a parent by the means with a new randomly generated numerical value, a new set is generated that partially inherits the properties of the parent. Value replacement means, generation method selection means for stochastically selecting the numerical value exchange and the new value replacement, and the group management using the elevator specification data, the traffic flow specification data and the new set in the simulation device. Evaluation means for obtaining an execution result when the algorithm is executed as a group management performance value; addition means for additionally storing only a good new set satisfying a predetermined addition condition in the storage means; satisfying a predetermined deletion condition Deleting means for deleting the inferior set from the storage means; and a plurality of sets stored and improved in the storage means. Extracting means for extracting the optimum set based on the group management performance value; and when any one of the elevator specification data and the traffic flow specification data changes, the search device re-searches the optimum set. An elevator group management system, comprising: re-search determination means for instructing. 36. A system for group-managing a plurality of elevators according to a group-management algorithm including a plurality of parameters, the apparatus controlling operation of a plurality of elevators installed in a building according to the group-management algorithm. , Elevator specification data representing the specifications of the plurality of elevators,
A group management device that outputs traffic flow specification data about passenger traffic in the building; andthe elevator management data and the traffic flow specification data output from the group management device; A search device that searches for an optimal set as a set that is a combination of parameter values to be given to the group management algorithm that has: a search device that stores a plurality of sets; A cross-type generating means for generating two new sets partially inheriting the characteristics of the parent by exchanging some numerical values between the two sets as the issued parents; Using the data, the traffic flow specification data and the new set, the execution result when the group management algorithm is executed Using the evaluation means to determine the value as a value, the addition of the new set to the storage means, and the deletion of the bad set from the storage means, and selecting the plurality of sets stored in the storage means to be superior. Means, extracting means for extracting an optimal set having the best group management performance value from a plurality of sets stored and improved in the storage means, and any one of the elevator specification data and the traffic flow specification data. An elevator group management system, comprising: re-search determination means for instructing the search device to re-search for the optimal set when a change has occurred. 37. A system for group-managing a plurality of elevators according to a group management algorithm including a plurality of parameters, the apparatus controlling operation of a plurality of elevators installed in a building according to the group management algorithm. , Elevator specification data representing the specifications of the plurality of elevators,
A group management device that outputs traffic flow specification data about passenger traffic in the building; andthe elevator management data and the traffic flow specification data output from the group management device; A search device that searches for an optimal set as a set that is a combination of parameter values to be given to the group management algorithm that has: a search device that stores a plurality of sets; Mutants that generate one new set that partially inherits the properties of the parent by replacing some parameter values in one set as the issued parent with new random numbers Generating means, using the elevator specification data, the traffic flow specification data and the new set, the group management algorithm Using the execution result when the is executed as the group management performance value, adding the new set to the storage means, and deleting the bad set from the storage means, Selecting means for improving a plurality of sets stored in the means, and extracting means for extracting an optimal set having the best group management performance value among the plurality of sets stored and improved in the storage means. An elevator group management, comprising: re-search determination means for instructing the search device to re-search the optimum set when any one of the elevator specification data and the traffic flow specification data changes. system.