CN110171753B

CN110171753B - Elevator dispatching strategy processing method, device, equipment and storage medium

Info

Publication number: CN110171753B
Application number: CN201910476087.2A
Authority: CN
Inventors: 李文海; 李良; 张永生; 章飞; 江荣钿
Original assignee: Hitachi Building Technology Guangzhou Co Ltd
Current assignee: Hitachi Building Technology Guangzhou Co Ltd
Priority date: 2019-06-03
Filing date: 2019-06-03
Publication date: 2021-09-21
Anticipated expiration: 2039-06-03
Also published as: CN110171753A

Abstract

The embodiment of the invention discloses a method, a device, equipment and a storage medium for processing an elevator dispatching strategy. The method comprises the steps of obtaining running state information of at least two elevators and external calling information of each floor; inputting the running state information and the external calling information into a preset strategy prediction model for strategy analysis to obtain an elevator initial scheduling strategy; constructing an initial strategy population based on the initial scheduling strategy; carrying out evolution treatment on the initial strategy population to obtain an evolution strategy population; according to the obtained elevator parameter information, determining the fitness of each candidate evolution strategy in the evolution strategy population; and taking the candidate evolution strategy with the fitness meeting the preset condition as a target scheduling strategy, and realizing the technical effect of making an adaptive target scheduling strategy for the complex calling situation.

Description

Elevator dispatching strategy processing method, device, equipment and storage medium

Technical Field

The embodiment of the invention relates to an elevator technology, in particular to an elevator dispatching strategy processing method, device, equipment and storage medium.

Background

Generally, the personnel waiting for the elevator can call the elevator through the external call button arranged on the elevator, and when the dispatching system of the elevator detects that the external call button on a certain floor is pressed, the elevator can be dispatched to run to the floor to carry the personnel waiting for the elevator. Furthermore, the building is equipped with a plurality of elevators carrying elevator waiting personnel, but higher requirements are provided for the intelligent degree of the scheduling system, and at least the requirements comprise reduction of elevator operation energy consumption, saving of average elevator waiting time of the elevator waiting personnel and the like. Generally, elevators are dispatched using an elevator dispatching algorithm.

However, elevator dispatching algorithms often need to consider many variables, such as internal call information, external call information, car weighing, running floors, etc. of an elevator. In the case of multivariate, it is often impossible to solve the optimal solution with a complete mathematical formula. In addition, even if the optimal solution can be obtained, the time taken for the optimal solution is too long, which causes poor real-time performance of the response of the dispatching system to the elevator calling. And the existing elevator dispatching algorithm can only be configured with limited dispatching strategies. Under the condition, when some elevator calling situations which are not configured by the elevator dispatching algorithm occur, the dispatching system can only adopt a similar dispatching strategy for handling, so that the dispatching of the elevator cannot obtain more reasonable elevator running energy consumption and shorter average elevator waiting time.

Disclosure of Invention

The invention provides an elevator dispatching strategy processing method, device, equipment and storage medium, which are used for adapting to complex elevator calling conditions and saving energy consumption of elevator dispatching.

In a first aspect, an embodiment of the present invention provides an elevator dispatching strategy processing method, including:

acquiring running state information of at least two elevators and external calling information of each floor;

inputting the running state information and the external calling information into a preset strategy prediction model for strategy analysis to obtain an elevator initial scheduling strategy;

constructing an initial strategy population based on the initial scheduling strategy;

carrying out evolution treatment on the initial strategy population to obtain an evolution strategy population;

according to the obtained elevator parameter information, determining the fitness of each candidate evolution strategy in the evolution strategy population;

and taking the candidate evolution strategy with the fitness meeting the preset condition as a target scheduling strategy.

Further, the operation state information includes: the number, the floor, the running direction, the weighing and the internal calling information of the elevator are obtained; the external calling information comprises uplink external calling information and downlink external calling information corresponding to each floor.

Further, before inputting the operation state information and the external calling information into a preset strategy prediction model for strategy analysis to obtain an elevator initial scheduling strategy, the method further includes:

acquiring sample data carrying policy marks, wherein the sample data comprises sample running state information of at least two elevators and sample external calling information of each floor, and the policy marks are provided with scheduling policies in an associated manner;

and carrying out model training by using the sample data to obtain a strategy prediction model.

Further, the step of inputting the operation state information and the external calling information into a preset strategy prediction model for strategy analysis to obtain an elevator initial scheduling strategy includes:

combining the running state information and the external calling information to obtain an elevator information vector;

inputting the elevator information vector into a preset strategy prediction model for strategy analysis to obtain a strategy mark output from the strategy prediction model;

and taking the dispatching strategy set in association with the strategy mark as the initial dispatching strategy of the elevator.

Further, the constructing an initial policy population based on the initial scheduling policy includes:

performing mutation processing on the initial scheduling strategy to obtain a mutation scheduling strategy;

and generating an initial strategy population comprising the variation scheduling strategy and the initial scheduling strategy.

Further, the evolving the initial strategy population to obtain an evolved strategy population includes:

determining the initial strategy population as an initialized candidate evolution strategy population;

selecting the intermediate scheduling strategy of the candidate evolution strategy population to obtain a selection strategy population;

performing cross processing on the intermediate scheduling strategies in the selection strategy population to obtain a cross strategy population;

carrying out variation processing on the intermediate scheduling strategy in the cross strategy population to obtain a variation strategy population;

determining the variation strategy population as the candidate evolution strategy population obtained by evolution;

and determining the candidate evolution strategy population meeting the evolution condition as the evolution strategy population, and taking the intermediate scheduling strategy in the evolution strategy population as the candidate scheduling strategy.

Further, the performing variation processing on the intermediate scheduling policy in the cross policy population to obtain a variation policy population includes:

determining the variation probability of each intermediate scheduling strategy in the cross strategy population;

and when the intermediate scheduling strategy is determined to be subjected to mutation processing according to the mutation probability, state information in the intermediate scheduling strategy is changed according to a preset rule, wherein the state information represents the response operation of each elevator to the external calling information of each floor.

Further, the changing the state information in the intermediate scheduling policy according to a preset rule includes:

when the state information of a first target floor in one elevator is changed from elevator calling response information to elevator calling neglect information, determining a first target elevator in an intermediate dispatching strategy, wherein the state information of the first target elevator on the first target floor is the elevator calling neglect information;

and changing the state information of the first target elevator at the first target floor into elevator calling response information.

Further, changing the state information in the intermediate scheduling policy according to a preset rule includes:

reading calling marks corresponding to all floors from the external calling information;

marking the calling elevator as an empty floor as a second target floor;

and changing the state information of each elevator in the intermediate dispatching strategy at the second target floor into neglecting elevator calling.

Further, the evolution condition includes at least one of the following:

the time condition of the evolution treatment, the time condition of the evolution treatment and the fitness condition of the optimal scheduling strategy;

the elevator dispatching strategy processing method further comprises the following steps:

and when the evolution condition is not met, continuously executing selection processing on the intermediate scheduling strategy of the candidate evolution strategy population to obtain a selection strategy population.

Further, the determining the fitness of each candidate evolution strategy in the evolution strategy population according to the obtained elevator parameter information includes:

obtaining elevator parameter information of the elevator;

according to the elevator parameter information, calculating an elevator energy consumption value, an average elevator waiting time and an average elevator taking time corresponding to each candidate dispatching strategy;

calculating the weighted sum of the energy consumption value, the average elevator waiting time and the average elevator taking time corresponding to each candidate scheduling strategy;

and taking the reciprocal of the weighted sum as the fitness of each candidate scheduling strategy.

Further, the acquiring elevator parameter information of the elevator comprises:

searching a parameter record corresponding to the number of the elevator in a database, wherein the weighing range of the parameter record is within the preset range of the current weighing of the elevator;

calculating the average value of each parameter in the parameter record;

combining the average values of the parameters into elevator parameter information of the elevator.

Further, the step of using the candidate evolution strategy with the fitness meeting the preset condition as a target scheduling strategy includes:

sorting the candidate scheduling strategies and the initial scheduling strategies from large to small according to fitness;

taking the scheduling strategy ranked first in the ranking result as a target scheduling strategy;

after the candidate evolution strategy with the fitness meeting the preset condition is taken as the target scheduling strategy, the method comprises the following steps:

dispatching the elevator using the target dispatch strategy in response to the outbound message.

Further, after the candidate evolution policy whose fitness meets the preset condition is taken as the target scheduling policy, the method further includes:

updating the policy prediction model using the target scheduling policy.

In a second aspect, an embodiment of the present invention further provides an elevator dispatching strategy processing apparatus, where the apparatus includes:

the information acquisition module is used for acquiring the running state information of at least two elevators and the calling information of each floor;

the initial scheduling strategy determining module is used for inputting the running state information and the external calling information into a preset strategy prediction model for strategy analysis to obtain an elevator initial scheduling strategy;

an initial strategy population constructing module, configured to construct an initial strategy population based on the initial scheduling strategy;

the evolution strategy population generating module is used for carrying out evolution treatment on the initial strategy population to obtain an evolution strategy population;

the fitness determining module is used for determining the fitness of each candidate evolution strategy in the evolution strategy population according to the obtained elevator parameter information;

and the target scheduling strategy determining module is used for taking the candidate evolution strategy with the fitness meeting the preset condition as a target scheduling strategy.

In a third aspect, an embodiment of the present invention further provides an elevator dispatching policy processing device, where the device includes: a memory and one or more processors;

the memory for storing one or more programs;

when executed by the one or more processors, cause the one or more processors to implement the elevator dispatch policy processing method of any of the first aspects.

In a fourth aspect, embodiments of the present invention also provide a storage medium containing computer-executable instructions, which when executed by a computer processor, are used to perform the elevator dispatching strategy processing method according to any one of the first aspect.

The embodiment of the invention obtains the running state information of at least two elevators and the external calling information of each floor; inputting the running state information and the external calling information into a preset strategy prediction model for strategy analysis to obtain an elevator initial scheduling strategy; constructing an initial strategy population based on the initial scheduling strategy; carrying out evolution treatment on the initial strategy population to obtain an evolution strategy population; according to the obtained elevator parameter information, determining the fitness of each candidate evolution strategy in the evolution strategy population; the candidate evolution strategy with the fitness meeting the preset condition is used as the target scheduling strategy, the problem that the elevator scheduling strategy configured in advance is limited and cannot be applied to the complex elevator calling condition is solved, and the technical effect of making the adaptive target scheduling strategy for the complex elevator calling condition is achieved.

Drawings

Fig. 1 is a flowchart of an elevator dispatching strategy processing method according to an embodiment of the present invention;

FIG. 2 is a flowchart of an evolution processing sub-method according to a second embodiment of the present invention;

fig. 3 is a flowchart of a fitness calculation sub-method according to a third embodiment of the present invention;

fig. 4 is a flowchart of an elevator dispatching strategy processing method according to the fourth embodiment of the present invention;

fig. 5 is a schematic structural diagram of an elevator dispatching strategy processing device according to a fifth embodiment of the present invention;

fig. 6 is a schematic structural diagram of an elevator dispatching strategy processing device according to a sixth embodiment of the present invention.

Detailed Description

The present invention will be described in further detail with reference to the accompanying drawings and examples. It is to be understood that the specific embodiments described herein are merely illustrative of the invention and are not limiting of the invention. It should be further noted that, for the convenience of description, only some of the structures related to the present invention are shown in the drawings, not all of the structures.

Example one

Fig. 1 is a flowchart of an elevator dispatching strategy processing method according to an embodiment of the present invention, which is applicable to elevator dispatching, and in particular, to dispatching of multiple elevators. The method may be performed by an elevator dispatching policy processing device.

Referring to fig. 1, the elevator dispatching strategy processing method specifically includes the following steps:

and S110, acquiring the running state information of at least two elevators and the calling information of each floor.

In this embodiment, each elevator is provided with an operation state information for indicating the state of the elevator in the operation process, and specifically, the operation state information may include: the number of the elevator, the floor, the running direction, the weighing, the internal calling information and the like. Wherein, the serial number of the elevator is used for uniquely identifying the elevator and is marked as N; the floor where the elevator is located is the floor where the current elevator runs and arrives, and is marked as F; the running direction is used for indicating that the elevator runs upwards or downwards and is marked as R, wherein R is 0 and represents an upward direction, and R is 1 and represents a downward direction; the weighing, denoted L, is the weight currently carried by the elevator and can be used to determine whether the elevator is in an overweight condition or whether the elevator can also carry passengers.

Further, the elevator calling information includes internal calling information and external calling information, and in this embodiment, the internal calling information is used as one of the operation state information of the elevator due to the condition of the internal calling request of each elevator corresponding to each internal calling information.

Calling in information

The call-in information is used to indicate the floor of the elevator car from which the passenger is going out. Generally, an elevator car is provided with an internal call button associated with each floor, and when it is detected that the internal call button is pressed, it indicates that a passenger needs to exit at the floor corresponding to the internal call button. The call-in information for each elevator at least includes: the case of an in-call request on each floor. For example, the calling-in information may be represented by a sequence, which is denoted as IC, specifically, each digit of the sequence IC represents a corresponding floor, and the value of each digit is used to represent the number of calling-in persons. If the numerical value is 0, no internal calling is indicated, if the numerical value is more than 0, the number of internal calling is indicated, and if the number of uncertain internal calling is 1, the number of internal calling is indicated. Further, the elevator dispatching strategy processing equipment can dispatch the elevator where the passenger is located to run to the floor corresponding to the internal calling button according to the internal calling information.

Second, external calling information

The hall call information is used for indicating the floor where the passenger who needs to take the elevator is located. Generally, each floor is provided with a waiting area, the waiting area is provided with an external calling button related to the floor where the passenger is located, and when the external calling button is detected to be pressed, the passenger indicates that the passenger needs to take the elevator on the floor where the external calling button is located. The calling-out information for each floor at least comprises the following information: the case of an outbound call on each floor. Further, the external call button includes an uplink external call button and a downlink external call button, which respectively correspond to the uplink external call request and the downlink external call request. When the uplink external calling button is pressed down, the fact that the floor where the uplink external calling button is located has an uplink external calling request is indicated, namely, a passenger needs to take an elevator to go up; when the down external calling button is pressed, the down external calling button indicates that the floor where the down external calling button is located has a down external calling request, namely that a passenger needs to take an elevator to go down. The corresponding external calling information comprises uplink external calling information and downlink external calling information corresponding to each floor. The same representation mode as the internal calling information, the uplink external calling information and the downlink external calling information can be respectively identified by a sequence and respectively marked as a sequence OCU and a sequence OCD. Illustratively, when a sequence of OCUs is used to represent the outbound upstream information, each digit of the sequence of OCUs represents a corresponding floor and the value of each digit represents the number of outbound upstream people. If the numerical value is 0, no uplink external calling is indicated, if the numerical value is more than 0, the number of the uplink external calling is indicated, and if the number of the uncertain uplink external calling is 1. Further, the elevator dispatching strategy processing equipment can dispatch the elevator to move to the floor corresponding to the external calling button to carry passengers according to the external calling information. Since there are a plurality of elevators, energy consumption values of elevators, average waiting times of passengers, and average boarding times of passengers, which are caused by scheduling of different elevators, are different. Therefore, the elevator dispatching strategy processing device is required to select a proper elevator dispatching strategy so as to optimize the energy consumption value of the elevator, the average waiting time of passengers and the average elevator taking time of the passengers.

And S120, inputting the running state information and the calling information into a preset strategy prediction model for strategy analysis to obtain an elevator initial scheduling strategy.

In this embodiment, the scheduling policy is used to indicate whether an elevator responds to a call request included in elevator call information, where the elevator call information includes external call information and internal call information, the external call information includes an external call request corresponding to each floor, the external call request includes an uplink external call request and a downlink external call request, and the internal call information includes an internal call request corresponding to the elevator and related to each floor. Specifically, the scheduling policy may be represented by a sequence P, where a value of each bit in the sequence P represents status information, and the status information includes elevator calling response information and elevator calling ignore information. If 1 is used for calling the elevator response information, 0 is used for calling the elevator neglect information. For example, there are m elevators in total, n floors, then P can be expressed as:

S₁₁,S₁₂,…,S_1n,S₂₁,S₂₂,…,S_2n,…,S_m1,S_m2,…,S_mn,X₁₁,X₁₂,…,X_1n,X₂₁,X₂₂,…,X_2n,…,X_m1,X_m2,…,X_mnwherein S is_xyWhether the elevator with the number of X responds to the Y-layer ascending external calling request or not is shown, 0 represents no response, namely calling elevator neglect information, 1 represents response, namely calling elevator response information, and X_xyAnd the elevator with the number of X responds to the Y-floor descending external calling request, 0 indicates no response, namely calling elevator neglect information, and 1 indicates response, namely calling elevator response information.

The policy prediction model in this embodiment may be used to preliminarily determine an elevator scheduling policy, i.e., an initial scheduling policy. Since the initial scheduling policy obtained by using the policy prediction model is limited, and cannot be applied to a complex calling request, and needs to be adjusted in actual use to obtain a target scheduling policy, this embodiment will further explain a specific adjustment manner.

In this embodiment, the policy preset model may be determined in two ways: elevator group control algorithm and machine learning algorithm.

Elevator group control algorithm

Conventional elevator group control algorithms may include: a First Come First served algorithm (FCFS), a Shortest Time to find floor First algorithm (SSTF), a SCAN algorithm (SCAN), etc.

1. First come first serve algorithm

The first-come first-serve algorithm is a random-serve algorithm, not only does not optimize the searched floor, but also has no real-time characteristic, and is the simplest elevator dispatching algorithm. It schedules according to the passenger's calling elevator request sequence. The algorithm has the advantages of fairness and simplicity, the elevator calling requests of all passengers can be processed in sequence, and the situation that the elevator calling requests of a certain passenger cannot be met for a long time can not occur. The performance of the method is acceptable under the environment with light load, but the performance of the algorithm is seriously reduced or even deteriorated under the condition with large load.

2. Shortest floor searching time priority algorithm

The shortest floor searching time priority algorithm emphasizes the optimization of the elevator for searching floors. The principle of the shortest time to find the floor for the priority algorithm to select the next service object is the shortest time to find the floor. Thus, the elevator calling request from the floor which can be reached first in the request queue is the next service object. In the case of heavy loads, the average response time of the shortest-distance floor-finding time-first algorithm is short, but the variance of the response time is large because some elevator-calling requests in the queue may not be responded for a long time, and a so-called "starvation" phenomenon occurs.

3. Scanning algorithm

The scanning algorithm is a sequential floor-by-floor service request that allows an elevator to travel continuously back and forth between the bottom most floor and the top most floor in response to elevator calls placed on floors in the same direction of elevator travel during the trip. The floor searching optimization is carried out, the efficiency is high, but the floor searching optimization is a non-real-time algorithm. The scanning algorithm solves the problem of elevator movement well, in the algorithm, each elevator responds to the elevator calling request of passengers, the order of the service of the passengers is determined by the distance between the position of the passenger sending the elevator calling request and the current elevator position, and the elevator calling requests of all the passengers in the same direction as the elevator running direction are completed in the process of one-time electricity upward running or downward running, so that the frequent back and forth movement of the elevator is avoided.

When the traditional elevator group control algorithm is used, preliminary strategy analysis can be carried out according to the acquired running state information corresponding to each elevator and the external calling information corresponding to each floor, so that an initial dispatching strategy is obtained. And each traditional elevator group control algorithm has respective advantages and disadvantages, and the initial dispatching strategy also needs to be adjusted to be applied to the elevator for actual elevator dispatching.

Two, machine learning algorithm

When a machine learning algorithm is used, the scheduling strategy can be set in a strategy mark association manner by acquiring sample data carrying strategy marks, wherein the sample data comprises sample running state information of at least two elevators and sample calling information of each floor; and carrying out model training by using the sample data to obtain a strategy prediction model. The strategy marks are strategy numbers, and each strategy mark is provided with a scheduling strategy in an associated mode. Further, the policy token carried by the sample data may be determined using any one of the above elevator group control algorithms and any combination thereof. That is, the strategy prediction model obtained by model training can replace any one of the elevator group control algorithms and any combination thereof. In one embodiment, the policy prediction model may be a neural network, a decision tree, a random forest, or the like. In this embodiment, a policy prediction model is taken as an example of a decision tree for explanation, so that efficiency and performance of model training can be improved.

Under the condition of using a machine learning algorithm, the running state information and the external calling information can be combined to obtain an elevator information vector; inputting the elevator information vector into a preset strategy prediction model for strategy analysis to obtain a strategy mark output from the strategy prediction model; and taking the dispatching strategy which is set in association with the strategy mark as the initial dispatching strategy of the elevator. In this embodiment, the combination mode of the operation state information and the external calling information is not limited, and only needs to be consistent with the combination mode used in the training process of the strategy prediction model. Illustratively, one way of representing the elevator information vector is:

OCU₁…OCU_n,OCD₁…OCD_n,N₁，F₁，R₁，L₁，IC₁₁…IC_1n，N₂，F₂，R₂，L₂，IC₂₁…IC_2n，…，N_m，F_m，R_m，L_m，IC_m1…IC_mnwherein n is the number of elevators; m is the number of floors; OCU₁…OCU_n，OCD₁…OCD_nIndicating external call information, OCU₁…OCU_nRepresenting uplink external calling information; OCD₁…OCD_nRepresenting downlink external calling information; n is a radical of_m，F_m，R_m，L_mRepresents N_mRepresenting the operating status information of the m-th elevator, N_mA number indicating an m-th elevator; f_mThe floor where the mth elevator is located; r₁Indicating the running direction of the m-th elevator; l is₁Representing the weighing of the m-th elevator; IC (integrated circuit)_m1…IC_mnIndicating recall information, wherein the IC_mnThis indicates that the m-th elevator has a call request for the n-th floor.

S130, constructing an initial strategy population based on the initial scheduling strategy.

Generally, the initial scheduling policy obtained by performing policy analysis by using the policy prediction model determined in the above two ways is generally limited and cannot correspond to all elevator calling situations. That is, when the operation state information and the external call information corresponding to the elevator are changed, the initial scheduling policy obtained by analysis cannot cope with the changed operation state information and the external call information because the policy prediction model does not correspond to all the operation state information and the external call information. Therefore, in order to overcome the problem, the scheduling policy may be used as a gene in a genetic algorithm, and further, the initial scheduling policy is evolved in a genetic algorithm manner, so as to finally obtain a target scheduling policy that is applicable to the occurrence of the change of the operating state information and the external calling information.

Specifically, in this embodiment, a variant scheduling policy may be obtained by performing a variant process on the initial scheduling policy; an initial strategy population is generated that includes a variant scheduling strategy and an initial scheduling strategy. The mutation processing is a processing mode of evolution processing in a genetic algorithm. Illustratively, when the number of scheduling policies in the initial policy population is set to N, then a mutation process is performed on the initial scheduling policies to obtain N-1 variant scheduling policies, so that the initial policy population includes: 1 initial scheduling strategy and N-1 variant scheduling strategies.

And S140, carrying out evolution treatment on the initial strategy population to obtain an evolution strategy population.

The evolutionary process in this embodiment adopts a genetic algorithm, where the evolutionary process at least includes one of the following: selection processing, crossover processing and mutation processing. Specifically, selection processing, cross processing and mutation processing are sequentially performed on the initial strategy population to obtain a candidate evolution strategy. The candidate evolution strategies constitute an evolution strategy population.

S150, according to the obtained elevator parameter information, the fitness of each candidate evolution strategy in the evolution strategy population is determined.

The elevator parameter information in this embodiment at least includes: energy consumption, power, distance, speed, time, etc. of the elevator in the various operating phases. Wherein, the motion stage is divided into the following steps according to the difference of the running direction and the acceleration: ascending acceleration, ascending constant speed, ascending deceleration, descending acceleration, descending constant speed and descending deceleration.

For example, elevator parameter information may be represented using table 1.

TABLE 1 Elevator parameter information Table

In this embodiment, specifically, the elevator energy consumption value, the average elevator waiting time, and the average elevator riding time corresponding to each candidate scheduling policy may be calculated according to the obtained elevator parameter information; further, the fitness of each candidate scheduling strategy is calculated according to the energy consumption value, the average elevator waiting time and the average elevator taking time. Wherein, the energy consumption value of the elevator refers to the energy consumed by the elevator when all elevators operate according to the candidate dispatching strategy; the average elevator waiting time refers to the average time from when the passenger presses an external calling button to when the passenger enters the elevator car when all elevators operate according to the candidate dispatching strategy; the average ride time refers to the average time a passenger presses the call-in button to reach the designated floor when all elevators are operating according to the candidate dispatch strategy.

The fitness in the embodiment comprehensively considers the energy consumption value, the average elevator waiting time and the average elevator taking time of the elevator. It should be noted that, when the evolution processing is performed on the initial strategy population, the fitness needs to be considered in each evolution processing, so that the candidate evolution strategies obtained by the evolution can meet the technical effects of less energy consumption value, shorter average waiting time and shorter average elevator taking time.

And S160, taking the candidate evolution strategy with the fitness meeting the preset condition as a target scheduling strategy.

In this embodiment, the candidate scheduling policy and the initial scheduling policy may be sorted from large to small according to the fitness; and taking the scheduling strategy ranked first in the ranking result as a target scheduling strategy, so that compared with the initial scheduling strategy, the determined target scheduling strategy optimizes the energy consumption value, the average elevator waiting time and the average elevator taking time.

According to the technical scheme of the embodiment, the running state information of at least two elevators and the calling information of each floor are obtained; inputting the running state information and the external calling information into a preset strategy prediction model for strategy analysis to obtain an elevator initial scheduling strategy; constructing an initial strategy population based on the initial scheduling strategy; carrying out evolution treatment on the initial strategy population to obtain an evolution strategy population; according to the obtained elevator parameter information, determining the fitness of each candidate evolution strategy in the evolution strategy population; the candidate evolution strategy with the fitness meeting the preset condition is used as the target scheduling strategy, the problem that the elevator scheduling strategy configured in advance is limited and cannot be applied to the complex elevator calling condition is solved, and the technical effect of making the adaptive target scheduling strategy for the complex elevator calling condition is achieved.

Example two

Fig. 2 is a flowchart of a sub-method of the evolution process according to the second embodiment of the present invention, and in this embodiment, on the basis of the above-mentioned embodiment, in the evolution process, consideration is added to the degree of use of each scheduling policy, where the degree of fitness is determined according to the energy consumption value, the average waiting time, and the average riding time generated after the elevator applies the scheduling policy, and step S140 is further refined into steps S1401 to S1407:

and S1401, determining the initial strategy population as an initialized candidate evolution strategy population.

In this embodiment, the initial policy population that is not subjected to the evolution process is used as the initialized candidate evolution policy population.

S1402, selecting the intermediate scheduling strategies of the candidate evolution strategy population to obtain a selection strategy population.

In this embodiment, the selection process is used to randomly select a part of the intermediate scheduling policies from the candidate evolution policy population, and construct a selection policy population including the selected part of the intermediate scheduling policies. Specifically, the probability of the intermediate scheduling policy being selected is related to the fitness of the intermediate scheduling policy.

Illustratively, the selection process may be performed by roulette, so that the probability of each intermediate scheduling policy being selected is proportional to the fitness thereof.

And S1403, performing cross processing on the intermediate scheduling strategies in the selection strategy population to obtain a cross strategy population.

In this embodiment, the cross processing is configured to randomly select two intermediate scheduling policies, randomly select a position of state information that needs to be cross processed, and interchange state information corresponding to the position in the two intermediate scheduling policies, where the state information includes, as described in the foregoing embodiment: the elevator calling system comprises elevator calling response information indicating response to the elevator calling request and elevator calling neglect information indicating non-response to the elevator calling request. Further, the position may be selected in one or more manners, and may be continuous or spaced, which is not limited in this embodiment.

The probability of the same intermediate scheduling strategy for cross processing can also be related to the fitness of the intermediate scheduling strategy, and the higher the fitness, the lower the probability of the intermediate scheduling strategy for cross processing is, so as to ensure that the intermediate scheduling strategy with high fitness can be continued to a new strategy population.

And S1404, carrying out mutation processing on the intermediate scheduling strategies in the cross strategy population to obtain a mutation strategy population.

In this embodiment, the mutation process is used to randomly select the intermediate scheduling policy and randomly select the position of the state information that needs to be mutated. Wherein the state information represents a response operation of each elevator to the hall call information of each floor. Specifically, the status information may include elevator calling response information indicating that an elevator calling request is responded, and elevator calling omission information indicating that an elevator calling request is not responded. Further, the state information corresponding to the position may be changed from the elevator calling response information to the elevator calling neglect information, or from the elevator calling neglect information to the elevator calling response information.

Specifically, for the application scenario of elevator scheduling, in this embodiment, it may be determined the variation probability of each intermediate scheduling policy in the cross policy population. The mutation probability can also be determined according to the fitness of the intermediate scheduling strategy, wherein the higher the fitness is, the smaller the mutation probability is. Further, when the intermediate scheduling strategy is determined to be mutated according to the mutation probability, the state information in the intermediate scheduling strategy is changed according to a preset rule.

It should be noted that the preset rule at least includes: and each external call is ensured to have at least 1 elevator response, if no external call exists on a specific floor, the state information corresponding to the specific floor in the intermediate scheduling strategy is fixed as elevator calling neglected information, and the external call and the current elevator running direction in the variation process are ensured to be the same as possible.

Specifically, this embodiment will exemplify the change of the state information in the intermediate scheduling policy according to the preset rule.

1) Ensuring that there are at least 1 elevator response per outbound call

Specifically, when the state information of a first target floor in one elevator is changed from the elevator calling response information to elevator calling neglect information during mutation processing, a first target elevator is determined in the intermediate dispatching strategy, and the state information of the first target elevator on the first target floor is the elevator calling neglect information; and changing the state information of the first target elevator at the first target floor into elevator calling response information.

2) If the specific floor does not have external calling, the state information corresponding to the specific floor in the intermediate scheduling strategy is fixed as calling neglected information

Reading calling marks corresponding to all floors from the external calling information; marking the calling elevator as an empty floor as a second target floor; and changing the state information of each elevator in the intermediate dispatching strategy at the second target floor into ignoring elevator calling.

S1405, determining the variation strategy population as the candidate evolution strategy population obtained by evolution.

In this embodiment, a variation strategy population obtained by performing evolution processing (including selection processing, crossover processing, and variation processing) is used as a candidate evolution strategy population obtained by evolution.

S1406, determining whether the candidate evolution policy population obtained by the evolution satisfies the evolution condition, if yes, performing step S1407, and if no, continuing to perform steps S1402-S1405 until the evolution condition is satisfied.

In this embodiment, the evolution condition at least includes one of the following conditions: the time condition of the evolution treatment, the time condition of the evolution treatment and the fitness condition of the optimal scheduling strategy.

First, the number of evolution process

In this embodiment, the steps corresponding to the selection process, the crossover process, and the mutation process are used as the steps of the evolution process, and further, the number of times of the evolution process is determined, and the steps including the selection process, the crossover process, and the mutation process are continuously executed once, and it is considered that only one time of the evolution process is executed. Further, when the number of times of the evolution processing is determined to exceed a preset number threshold, the number condition of the times of the evolution processing is determined to be met.

Second, time condition of evolution treatment

In this embodiment, similarly, the steps corresponding to the selection process, the crossover process, and the mutation process are used as the steps of the evolution process, and further, the execution time of the evolution process is determined, and the steps including the selection process, the crossover process, and the mutation process are continuously executed once, and it is considered that only one time of the evolution process is executed. Further, when the execution time of the evolution process is determined to exceed the preset time threshold, it is determined that the time condition of the evolution process is satisfied.

Thirdly, the fitness condition of the optimal scheduling strategy.

In this embodiment, after each evolution process is completed, the fitness of each intermediate scheduling policy in the candidate evolution policy population is calculated, and when it is determined that the fitness exceeds a preset fitness threshold (usually, the fitness of the initial scheduling policy is used as the fitness threshold), it is determined that the fitness condition of the optimal scheduling policy is satisfied.

S1407, determining the candidate evolution strategy population meeting the evolution condition as the evolution strategy population, and taking the intermediate scheduling strategy in the evolution strategy population as the candidate scheduling strategy.

EXAMPLE III

Fig. 3 is a flowchart of a fitness calculating sub-method according to a third embodiment of the present invention, and this embodiment adds details of calculating the fitness by using the elevator parameter information to calculate the energy consumption value, the average waiting time, and the average boarding time, and adding details of calculating the fitness by using the energy consumption value, the average waiting time, and the average boarding time on the basis of the third embodiment. Specifically, step S150 in the above embodiment may be further refined into steps S1501 to S1504:

s1501, obtaining elevator parameter information of the elevator.

In this embodiment, the elevator parameter information will be described by taking as an example that the elevator parameter information includes at least the parameter information shown in table 1.

In the embodiment, the weighing range of the parameter record is within the preset range of the current weighing of the elevator by searching the parameter record corresponding to the serial number of the elevator in the database; calculating the average value of each parameter in the parameter record; the average values of the parameters are combined to elevator parameter information of the elevator. Illustratively, the database stores parameter records of a preset period (such as 3 months), and the parameter records are elevator parameter information marked with acquisition time. Further, when elevator parameter information of the elevator needs to be acquired, a parameter record which corresponds to the number N of the elevator and is within a weighing range (such as (1 +/-5%) × L) can be searched from the database according to the number N of the elevator and the current weighing L, and then an average value of each parameter in the searched parameter record is calculated; the average values of the parameters are combined to elevator parameter information of the elevator. If the parameter records corresponding to the elevator with the number N and within the weighing range (such as (1 +/-5%) × L) are obtained, further, the uplink acceleration energy consumption in all the parameter records is averaged, and then the average value is used as the uplink acceleration energy consumption in the obtained elevator parameter information. It should be noted that when no parameter record can be found in the weighing range (e.g., (1 + -5%). times.L), then the record with the smallest absolute value difference from the weighing L is used.

In this embodiment, how to collect the parameter record in real time and record the parameter record in the database is not limited, and this embodiment will be described by way of example.

In one embodiment, elevator operation data is collected in real time while the elevator is operating, the elevator operation data including at least one of: the running direction of the elevator, the current time of the elevator, the weighing of the elevator, the voltage and current of a traction system and the running distance of the elevator are calculated, and further, a parameter record is obtained through the running data of the elevator, namely, the parameter information of the elevator marked with the acquisition time.

Generally, an elevator operation route is divided into an acceleration section, a constant speed section and a deceleration section, the basis for judging the three sections is speed v and acceleration a, v >0 and a >0 are acceleration sections, v >0 and a ═ 0 are constant speed sections, v >0 and a <0 are deceleration sections, and v is equal to 0 and stops.

Further, the plurality of continuous elevator operation data collected in the elevator one-time operation process (stop → acceleration stage → uniform stage → deceleration stage → stop) may be recorded in a data collection sequence, and exemplarily may include a time sequence T, a voltage sequence U, a current sequence I, and a distance sequence D, where the time sequence T is a sequence including the current time of a plurality of continuous elevators, and the voltage sequence U is a sequence including a plurality of continuous traction system voltages. Exemplarily, it can be written as:

time series T ═ T₀，t₁，t₂，…，t_n]

Voltage sequence U ═ U₀，u₁，u₂，…，u_n]

Current sequence I ═ I₀，i₁，i₂，…，i_n]

Distance sequence D ═ D₀，d₁，d₂，…，d_n]

Further, the time increment sequence DT, the distance increment sequence DD, the velocity sequence V, the velocity increment sequence DV, the acceleration sequence a, and the like can be calculated through the time sequence T and the distance sequence D. Exemplarily, it can be written as:

time increment sequence DT ═ 0，dt₁，dt₂，…，dt_n]＝[0，…，dt_ac，dt_ac+1，…，dt_cs，dt_cs+1，…，dt_n]

Distance increment sequence DD ═ 0, DD₁，dd₂，…，dd_n]＝[0，…，dd_ac，dd_ac+1，…，dd_cs，dd_cs+1，…，dd_n]

Velocity sequence V ═ 0, V₁，v₂，…，v_n]＝[0，…，v_ac，v_ac+1，…，v_cs，v_cs+1，…，v_n]

Speed increment sequence DV ═ 0, DV₁，dv₂，…，dv_n]＝[0，…，dv_ac，dv_ac+1，…，dv_cs，dv_cs+1，…，dv_n]

Acceleration sequence a ═ 0, a₁，a₂，…，a_n]＝[0，…，a_ac，a_ac+1，…，a_cs，a_cs+1，…，a_n]

In this embodiment, the switching time point of each segment can be determined according to the discrimination criteria of the acceleration segment, the uniform velocity segment and the deceleration segment, and the time point of the acceleration segment converted into the uniform velocity segment is set as t_acThe time point of the uniform velocity segment converted into deceleration is t_csThe membership segment of each data in the acquired data sequence may be determined.

Further, the acceleration energy consumption E can be performed using the above time increment series DT, distance increment series DD, velocity series V, velocity increment series DV, and acceleration series a_aEnergy consumption of deceleration E_sConstant power P_cAcceleration distance D_aDeceleration distance D_sConstant speed rated speed V_cAcceleration time T_aTime of deceleration T_sAnd (4) calculating. It should be noted that the energy consumption E is accelerated_aEnergy consumption of deceleration E_sConstant power P_cAcceleration distance D_aDeceleration distance D_sConstant speed rated speed V_cAnd acceleration of the motorTime T_aTime of deceleration T_sThe calculation of these parameters requires consideration of whether the elevator is in the up-going phase or the down-going phase, as detailed in table 1. Of course, the determination can be made in the following calculation manner in both the uplink phase and the downlink phase.

1. Accelerated energy consumption E_a

Traction system voltage sequence U through acceleration section_acCurrent sequence I_acAnd a sequence of time increments DT_acCalculated, the formula is as follows:

2. energy consumption of deceleration E_s

Traction system voltage sequence U through speed reduction section_csCurrent sequence I_csAnd a sequence of time increments DT_csCalculated, the formula is as follows:

3. constant power P_c

By means of uniform segment voltage sequences U_cCurrent sequence I_cThe average value is calculated, and the formula is as follows:

4. acceleration distance D_a

Distance increment sequence DD by acceleration segment_acThe formula is as follows:

5. deceleration distance D_s

By a sequence of deceleration segment distances DD_csCalculation of the formulaThe following:

6. constant rated speed V_c

The speed sequence V of the uniform speed segment is calculated, and the formula is as follows:

7. acceleration time T_a

By accelerating the time series DT of the segments_acThe formula is as follows:

8. deceleration time T_s

By decelerating the time series DT of the segments_csThe formula is as follows:

wherein M is used for determining the sequence range corresponding to each operation stage, such as U is needed for the voltage sequence U₀，u₁，u₂，…，u_M。

Further, the parameter record at least comprises: collecting the number S, the number N of the elevator, the running direction R of the elevator, the weighing L of the elevator and the acceleration energy consumption E_aReduction energy consumption E_sConstant power P_cAcceleration distance D_aDistance of deceleration D_sConstant velocity rated velocity V_cAcceleration time T_aTime of deceleration T_sAnd acquisition time CT. Wherein, the acquisition number is automatically generated in an increasing mode; the running direction R and the weighing L of the elevator can be collected by the original elevator system or adjusted by the elevator in the embodimentAcquiring by using degree strategy processing equipment; the elevators in the embodiment belong to an elevator group, and the number N of the elevator is the number of the elevator in the elevator group; the acquisition time CT can also be generated automatically, and the system time can be directly adopted.

S1502, according to the elevator parameter information, calculating an elevator energy consumption value, an average elevator waiting time and an average elevator taking time corresponding to each candidate dispatching strategy.

In this embodiment, the energy consumption value, the average elevator waiting time, and the average elevator riding time are three major criteria for evaluating elevator scheduling policies, and may be used to calculate the fitness of each scheduling policy. And further, evolving a scheduling strategy according to the fitness. Generally, the more the scheduling policy meets the three criteria, the higher the fitness.

In this embodiment, three standards need to be quantized to be executed in the computer. Correspondingly, the elevator dispatching strategy is evaluated by adopting the following method:

1) and (4) a minimum energy consumption evaluation method.

2) And (4) a shortest average waiting time evaluation method.

3) And (4) a shortest average elevator taking time evaluation method.

Wherein, the lowest energy consumption means that the energy (power) consumed by all elevators in one-time complete dispatching strategy is the lowest; the shortest average waiting time refers to the shortest average waiting time from the time when a passenger presses an external calling button to the time when the passenger enters the elevator car; the shortest average elevator taking time means that the average elevator taking time from the time when the passenger presses the calling-in button to the time when the passenger arrives at the appointed floor is shortest. For convenience of discussion, the present embodiment assigns a function sign to each evaluation method:

energy consumption evaluation function: e_e

Average waiting time evaluation function: t is_w

Average elevator ride time evaluation function: t is_l

Wherein, the energy consumption evaluation function E_eUsed for calculating the energy consumption value and the average waiting time evaluation function T_wUsed for calculating average elevator waiting time and evaluating function T of average elevator taking time_lFor calculating the average boarding time.

In this embodiment, the manner of calculating the energy consumption value, the average waiting time, and the average riding time will be described by way of example.

Energy consumption value of elevator

The elevator energy consumption is composed of energy consumption of a plurality of subsystems, such as an air conditioning system, a lighting system, a door system, a traction system and the like. Because the energy consumption of the traction system is the main energy consumption of the elevator and has close relation with the dispatching strategy, the lowest energy consumption evaluation function E_eAnd calculating by adopting the energy consumption of the traction system.

1. Elevator energy consumption E of a complete operating process_rComputing

In this embodiment, the elevator once complete operation process includes: door closing → starting → stopping → door opening, then the elevator energy consumption E_rCalculating the formula:

E_r＝W_c+E_p+W_o＝E_p+C_d

wherein, W_cEnergy consumption for closing the door, W_oIs energy consumption for opening the door due to W_cAnd W_oIs substantially fixed and can therefore be considered to be a constant C_d，C_d＝W_c+W_o，E_pRefers to the energy consumption of a traction system.

Furthermore, the energy consumption E of the traction system_pMainly by accelerating energy consumption E_aUniform energy consumption and deceleration energy consumption E_sHowever, in some cases, the elevator only goes through an acceleration section and a deceleration section because of a short travel distance, and calculation needs to be distinguished.

And if the elevator runs from the floor m to the floor n in one complete running mode and the distance is l, then:

wherein: e_aIs accelerated energy consumption, E_sIs the deceleration of the energy consumption, D_aIs the acceleration distance, D_sIs the deceleration distance, P is the uniform power (here, the uniform power P)_cSame), V is the constant nominal speed (this isAt and at a constant speed of rated speed V_cSame), C)_dIs the door system energy consumption, l is the travel distance. Note that E_a、E_s、D_a、D_sP, V is the elevator parameter information obtained, and distinguishes the up-down going when m<n is an uplink, m>And n is descending.

2. Energy consumption value E_eComputing

Generally, an elevator dispatching strategy will result in multiple elevator start-stops, i.e. including multiple complete runs. That is, in calculating the energy consumption value E_eIn time, the elevator energy consumption E of a plurality of complete operation processes needs to be comprehensively considered_r。

1) Counting of elevator start and stop times

Generally, the number of times of starting and stopping an elevator is determined by an elevator dispatching strategy, the current floor and the running direction of the elevator and an internal calling floor determined according to internal calling information.

Suppose that in the current scheduling strategy, the uplink external call information of a certain elevator in the elevator group is OCU ═ U₀,U₁,…,U_n]Wherein the subscript of U denotes the floor, U_nThe number of the calling persons in the upper row corresponding to the nth floor is shown, and the number of the calling persons in the upper row is not determined and is shown as 1; the downlink external call information is OCD ═ D₀,D₁,…,D_m]Wherein the subscript of D denotes the floor, D_mThe number of the calling persons in the lower row corresponding to the mth floor is shown, and the number of the calling persons in the lower row is not determined and is shown as 1; the corresponding calling information of the current elevator is IC ═ I₀,I₁,…,I_k]Wherein the subscript of I denotes the floor, I_kThe number of calling persons in the kth floor is shown, and the number of calling persons in the uncertain floor is shown as 1; the floor where the elevator is located is F; the current running direction of the elevator is R, wherein R ═ 0 denotes an upward direction, and R ═ 1 denotes a downward direction, then:

where Round refers to the number of times of starting and stopping, Size () refers to the number of elements not 0 in the statistical sequence, U refers to a union, e.g., uodu DRepresents the union of sequence O and sequence D. Suppose that the intra-summoning and the external-summoning are in the same direction and gather the sequence ICOC ═ C₀,C₁,…,C_n]The calculation method is as follows: when the direction is uplink, the direction corresponds to IC [. sup.C.sub.OCU ]; when the direction is descending, the direction corresponds to IC [. sup.C.sub.D ]; reverse direction external recall sequence OC ═ OC₀,OC₁,…,OC_m]Wherein the reverse direction refers to the direction opposite to the direction of elevator travel.

2) Energy consumption E in connection with an elevator₀Is calculated by

An energy consumption calculation formula of an elevator:

wherein F, c0, cn, oc0 and oc all represent floors, and n and m both represent the number of floors; e_oMeans energy consumption of a certain elevator in the group in the current scheduling strategy, E_r(F,c0)Means that the elevator is started from F floor to ICOC c0]Energy consumption of a complete run with layer stop, E_r(c,c-1)Means that the elevator is driven by ICOC [ c-1]Layer Start to ICOC [ c ]]Energy consumption of a complete run with layer stop, E_r(cn,oc0)Means that the elevator is driven by ICOC [ cn [)]Layer startup to OC [ OC0]Energy consumption for a complete run through vehicles with layer stop, E_r(oc,oc-1)Means that the elevator is driven by OC [ OC-1 ]]Layer initiation to OC [ OC ]]Energy consumption of a complete run with layer stop, E_rCalculating the energy consumption E of the elevator in the complete running process_rAnd (4) calculating mode.

3) Energy consumption value E corresponding to all elevators in elevator group_eIs calculated by

The total energy consumption of the dispatching strategy is obtained by summing the energy consumption of all the current dispatching strategies of the elevators in the elevator group, and is represented as follows:

wherein m denotes the number of elevators, E_oiRepresenting the energy consumption of the i-th elevator applying the current scheduling policy.

Second, average waiting time

The average elevator waiting time is obtained by summing and averaging the elevator waiting times of all the elevators in the elevator group in all the external calling requests of the current dispatching strategy. Because the elevator responds to the internal calling request and the external calling request simultaneously in the running process, the elevator waiting time of a certain external calling request needs to be calculated while considering the complete running process of the internal calling request and the complete running process of the external calling request.

1. Waiting time T about calling floor m out in the same direction of current elevator_smIs calculated by

When the current scheduling strategy is applied, calling the elevator waiting time T of the floor m outwards in the same direction_smIt can be expressed as:

wherein, T_wr(F,i0)Means that the elevator is started from F floor to ICOC [ i0 ]]Time of one complete run of the course of the layer stop, T_lr(i-1, i) means that the elevator is driven from IC [ i-1 ]]Layer initiation to IC [ i ]]Time of one complete run of the course of the layer stop, T_wr(n,m)Means that the slave elevator is slave ICOC [ k]Layer Start to ICOC [ m ]]Time for one complete run of the process with the layer stopped.

2. Total waiting time T in the same direction for current elevator_sIs calculated by

Setting the same direction external calling floor sequence OS ═ OS₀,OS₁,…,OS_n]If the current elevator adopts the same-direction total elevator waiting time T of the current scheduling strategy_sIt can be expressed as:

wherein, T_sm(i)Is referred to as OS [ i ]]External calling of the floor for the time of waiting, T_smThe calculations are as described above.

3. Waiting time T for calling floor m to the opposite direction of current elevator_omIs calculated by

Because the elevator can respond to the external calling request in the opposite direction only after finishing responding to the external calling request and the internal calling request in the same direction, the elevator waits for the elevator time T for the floor m in the opposite direction_omIt can be expressed as:

wherein, T_sIs the total waiting time in the same direction, T_lr(i-1,i)Means that the elevator is driven by IC [ i-1 ]]Layer initiation to IC [ i ]]Time of one complete run of the course of the layer stop, T_wr(n,m)Means that the elevator is driven by IC k]Layer start to OC [ m ]]Time for one complete run of the process with the layer stopped.

4. Total waiting time T in reverse direction for current elevator_oIs calculated by

Setting a reverse direction external floor calling sequence OO ═ OO₀,OO₁,…,OO_n]If the current scheduling strategy is applied, the total elevator waiting time T in the reverse direction of the current elevator_oIt can be expressed as:

wherein, T_om(i)Means that the first floor OO [ i ] is called outwards in the reverse direction]Time of waiting for stairs, T, of the floor_omThe calculations are as described above.

5. Calculation of average waiting time for n elevators in elevator group

In conclusion, the average elevator waiting time T of the current dispatching strategies of n elevators in the elevator group_wIt can be expressed as:

wherein, T_s(i)Is the total waiting time T of the ith elevator in the same direction_o(i)Is the total waiting time, size (OS) in the reverse direction of the ith elevator_(i)) Refers to the number of floors called out by the ith elevator in the same direction,size(OO_(i)) Is the number of floors called out in the opposite direction of the ith elevator.

Third, average elevator taking time

The average elevator taking time is obtained by averaging the elevator taking times corresponding to all internal calling requests in the current dispatching strategy of all elevators in the elevator group. Generally, in response to an internal call request, when a passenger determines that the traveling direction of an arriving elevator is the same as the direction of his/her own external call request, the passenger enters the elevator, presses a destination floor to which he/she goes, and the elevator receives the internal call request with respect to the destination floor. In addition, during the operation of the elevator, the response to the external call request in the same direction may be required, so the elevator taking time should calculate the internal call response time and the external call response time.

1. Elevator time T for calling floor m in response to current elevator_lmIs calculated by

When the current elevator is in the current dispatching strategy, responding to the elevator taking time T of the internal calling floor m_lmIt can be expressed as:

wherein, T_lr(F,c0)Means that the elevator is started from the current floor F to ICOC [ c0 ]]Time of one complete run of the course of the layer stop, T_wr(c-1,c)Means that the elevator is driven by ICOC [ c-1]Layer Start to ICOC [ c ]]Time of one complete run of the course of the layer stop, T_lr(k,m)Means that the elevator is driven by ICOC k]Starting to ICOC [ m ]]Time for one complete run of the process with the layer stopped.

2. Total time to take elevator T for current elevator_ltIs calculated by

Setting the internal calling information of the current elevator as IC ═ I₀,I₁,…,I_k]If the current dispatching strategy is adopted, the total elevator waiting time T of the current elevator_ltIt can be expressed as:

wherein, T_lm(i)Means responding to the internal calling of the IC i]Time to take the floor, T_lmThe calculations are as described above.

3. Average elevator taking time T for n elevators in elevator group_lIs calculated by

Average elevator riding time T of n elevators in elevator group when applying current dispatching strategy_lIt can be expressed as:

wherein, T_lt(i)Means the total time of taking the elevator, size (IC) of the i-th elevator_(i)) Which refers to the number of floors called in the i-th elevator.

It should be noted that in the above embodiment, the passenger waiting time and riding time are related to the elevator car operating time and the door system opening and closing time, and therefore, the passenger waiting time T is set to be equal to the elevator car operating time_wrAnd the elevator taking time T_lrCalculating the formula:

T_wr/＝T_c+T_v+T_o＝T_v+T_d

the running time of the elevator car mainly comprises acceleration time, constant speed time and deceleration time, but in some cases, the elevator only passes through the acceleration section and the deceleration section due to the fact that the running distance is short, and calculation needs to be distinguished.

Wherein, T_cTime of closing the door, T_oIs the time of opening the door due to T_cAnd T_oIs substantially fixed, and can therefore be considered to be a constant T_dCan use T_d＝T_c+T_oIndicating the door system open and close times. T is_vThe calculation can be performed in the following manner:

if the elevator runs completely from the floor m to the floor n (the elevator is currently at the floor m, the elevator waiting time is calculated when the user calls the elevator at the floor n, and the elevator taking time is calculated when the user calls the elevator from the floor m to the floor n), and the distance is l:

wherein, T_aIs the time, T, of the acceleration period of the elevator_sTime of elevator deceleration section, D_aIs the acceleration distance, D, of the elevator_sIs the deceleration distance of the elevator, V is the rated speed of the uniform speed section of the elevator (here, the rated speed V is equal to the constant speed_cSame) T_dIs the door opening and closing time of the door system. Note that T_a、T_s、D_a、D_sV is the obtained elevator parameter information, and distinguishes the up-down going when m<n is an uplink, m>And n is descending.

And S1503, calculating the weighted sum of the energy consumption value, the average elevator waiting time and the average elevator taking time corresponding to each candidate scheduling strategy.

In this embodiment, in the implementation process, the energy consumption value, the average elevator waiting time, and the average elevator taking time are mutually affected, and the technical effects of the lowest energy consumption value, the shortest average elevator waiting time, and the shortest average elevator taking time cannot be simultaneously obtained, so that the three standards of the energy consumption value, the average elevator waiting time, and the average elevator taking time need to be adjusted in a balanced manner.

In this embodiment, the three criteria are balanced and adjusted by the comprehensive evaluation function, which is recorded as:

F＝a*E_e+b*T_w+c*T_l

wherein a, b and c are weighting factors, and can be adjusted according to application scenarios, a is greater than or equal to 0 and less than or equal to 1, b is greater than or equal to 0 and less than or equal to 1, c is greater than or equal to 0 and less than or equal to 1, and a + b + c is equal to 1. In this embodiment, the weighted sum of the energy consumption value, the average waiting time, and the average riding time corresponding to each candidate scheduling policy is calculated, that is, a weighting factor is added, so as to more flexibly synthesize 3 evaluation functions, that is, balance adjustment three criteria. Meanwhile, a certain evaluation function may be directly used as a criterion, and a smaller evaluation function value indicates a better corresponding scheduling policy, and the fitness is higher.

For example:

F＝0.2*E_e+0.5*T_w+0.3*T_l

the lowest energy consumption is represented to be 0.2 weight, the average elevator waiting time is 0.5 weight, the average elevator taking time is 0.3 weight, and the final comprehensive evaluation function result is biased to the average elevator waiting time.

For another example:

F＝0*E_e+1.0*T_w+0*T_l

the average waiting time is evaluated without considering the energy consumption value and the average elevator taking time.

S1504, taking the reciprocal of the weighted sum as the fitness of each candidate scheduling strategy.

In this embodiment, since a smaller weighted sum indicates a better corresponding scheduling policy, the higher the fitness, the better the corresponding scheduling policy may be indicated by using the inverse of the weighted sum as the fitness. Furthermore, since the energy consumption value, the average elevator waiting time and the average elevator taking time are weighted and comprehensively considered, the energy consumption value, the average elevator waiting time and the average elevator taking time are also considered when the fitness of the scheduling strategy is evaluated.

Example four

Fig. 4 is a flowchart of an elevator dispatching strategy processing method according to a fourth embodiment of the present invention, which is further detailed based on the foregoing embodiment, and referring to fig. 4, the elevator dispatching strategy processing method specifically includes the following steps:

s410, obtaining the running state information of at least two elevators and the calling information of each floor.

And S420, inputting the running state information and the calling information into a preset strategy prediction model for strategy analysis to obtain an elevator initial scheduling strategy.

The policy prediction model in this embodiment may be used to preliminarily determine an elevator scheduling policy, i.e., an initial scheduling policy. The initial scheduling strategy obtained by using the strategy prediction model for analysis is limited, so that the method cannot be applied to complex elevator calling requests, and the target scheduling strategy is obtained by adjusting the strategy during actual use.

S430, constructing an initial strategy population based on the initial scheduling strategy.

In this embodiment, a variant scheduling policy is obtained by performing a variant process on the initial scheduling policy; and generating an initial strategy population comprising the variation scheduling strategy and the initial scheduling strategy.

S440, carrying out evolution treatment on the initial strategy population to obtain an evolution strategy population.

S450, according to the obtained elevator parameter information, the fitness of each candidate evolution strategy in the evolution strategy population is determined.

In this embodiment, specifically, the elevator energy consumption value, the average elevator waiting time, and the average elevator riding time corresponding to each candidate scheduling policy may be calculated according to the obtained elevator parameter information; further, the fitness of each candidate scheduling strategy is calculated according to the energy consumption value, the average elevator waiting time and the average elevator taking time.

1) and (4) a minimum energy consumption evaluation method.

2) And (4) a shortest average waiting time evaluation method.

3) And (4) a shortest average elevator taking time evaluation method.

energy consumption evaluation function: e_e

Average waiting time evaluation function: t is_w

Average elevator ride time evaluation function: t is_l

Further, in this embodiment, in the implementation process, the energy consumption value, the average elevator waiting time, and the average elevator taking time are mutually affected, and the technical effects of the lowest energy consumption value, the shortest average elevator waiting time, and the shortest average elevator taking time cannot be simultaneously obtained, so that the three criteria of the energy consumption value, the average elevator waiting time, and the average elevator taking time need to be adjusted in a balanced manner.

F＝a*E_e+b*T_w+c*T_l

And S460, taking the candidate evolution strategy with the fitness meeting the preset condition as a target scheduling strategy.

S470, using the target dispatching strategy to dispatch the elevator so as to respond to the calling-out information.

In this embodiment, a response operation to the external call information may be acquired from the target scheduling policy. Illustratively, there are m elevators in total, n floors, and assuming a target dispatch policy of P, it can be expressed as:

S₁₁,S₁₂,…,S_1n,S₂₁,S₂₂,…,S_2n,…,S_m1,S_m2,…,S_mn,X₁₁,X₁₂,…,X_1n,X₂₁,X₂₂,…,X_2n,…,X_m1,X_m2,…,X_mnwherein S is_xyDisplay weaveWhether the elevator with the number X responds to the Y-layer ascending external calling request or not, wherein 0 represents no response, namely, the elevator calling ignores information; 1 represents response, namely elevator calling response information, and the elevator with the dispatching number X runs to the Y floor to respond to the ascending external calling request of the Y floor; x_xyWhether the elevator with the number of X responds to the Y-layer descending external calling request or not is shown, and 0 indicates no response, namely, the elevator calling ignores the information; 1 represents response, namely elevator calling response information, and the elevator with the dispatching number X runs to the Y floor to respond to the downlink external calling request of the Y floor.

And S480, updating the strategy prediction model by using the target scheduling strategy.

In this embodiment, when the fitness of the target scheduling policy is higher than that of the initial scheduling policy, the current operation state information and the external calling information may be used as sample data, and a policy tag associated with the target scheduling policy may be added to the sample data. And updating the policy prediction model using the sample data. Illustratively, model training is performed on the strategy prediction model by using the sample data, so as to obtain an updated strategy prediction model. The strategy marks are strategy numbers, and each strategy mark is provided with a scheduling strategy in an associated mode. In one embodiment, the policy prediction model may be a neural network, a decision tree, a random forest, or the like. In this embodiment, a policy prediction model is taken as an example of a decision tree for explanation, so that efficiency and performance of model training can be improved.

EXAMPLE five

Fig. 5 is a schematic structural diagram of an elevator dispatching strategy processing device according to a fifth embodiment of the present invention, which is applicable to elevator dispatching and the method can be integrated into an elevator dispatching strategy processing device.

Referring to fig. 5, the elevator dispatching strategy processing device specifically comprises the following structures: an information acquisition module 510, an initial scheduling policy determination module 520, an initial policy population construction module 530, an evolution policy population generation module 540, a fitness determination module 550, and a target scheduling policy determination module 560.

And an information obtaining module 510, configured to obtain operation state information of at least two elevators and external call information of each floor.

And an initial scheduling policy determining module 520, configured to input the operation state information and the external call information into a preset policy prediction model for policy analysis, so as to obtain an elevator initial scheduling policy.

An initial policy population constructing module 530 configured to construct an initial policy population based on the initial scheduling policy.

And the evolution strategy population generating module 540 is configured to perform evolution processing on the initial strategy population to obtain an evolution strategy population.

And an adaptability determination module 550, configured to determine adaptability of each candidate evolution policy in the evolution policy population according to the obtained elevator parameter information.

And a target scheduling policy determining module 560, configured to use the candidate evolution policy whose fitness meets a preset condition as a target scheduling policy.

On the basis of the above technical solution, the operation state information includes: the number, the floor, the running direction, the weighing and the internal calling information of the elevator are obtained; the external calling information comprises uplink external calling information and downlink external calling information corresponding to each floor.

On the basis of the above technical solution, the apparatus further includes:

and the sample data acquisition module is used for acquiring sample data carrying strategy marks before inputting the running state information and the external calling information into a preset strategy prediction model for strategy analysis to obtain an elevator initial dispatching strategy, wherein the sample data comprises sample running state information of at least two elevators and sample external calling information of each floor, and the strategy marks are provided with dispatching strategies in an associated manner.

And the model training module is used for performing model training by using the sample data to obtain a strategy prediction model.

On the basis of the above technical solution, the initial scheduling policy determining module 520 includes:

and the elevator information vector acquisition unit is used for combining the running state information and the external calling information to obtain an elevator information vector.

And the strategy analysis unit is used for inputting the elevator information vector into a preset strategy prediction model for strategy analysis to obtain a strategy mark output from the strategy prediction model.

And the initial dispatching strategy determining unit is used for taking the dispatching strategy which is set in association with the strategy mark as the initial dispatching strategy of the elevator.

On the basis of the above technical solution, the initial policy population constructing module 530 includes:

and the variation scheduling strategy obtaining unit is used for executing variation processing on the initial scheduling strategy to obtain a variation scheduling strategy.

And the initial strategy population constructing unit is used for generating an initial strategy population comprising the variation scheduling strategy and the initial scheduling strategy.

On the basis of the above technical solution, the evolution strategy population generating module 540 includes:

and the candidate evolution strategy population initializing unit is used for determining the initial strategy population as an initialized candidate evolution strategy population.

And the selection processing unit is used for carrying out selection processing on the intermediate scheduling strategy of the candidate evolution strategy population to obtain a selection strategy population.

And the cross processing unit is used for carrying out cross processing on the intermediate scheduling strategies in the selection strategy population to obtain a cross strategy population.

And the variation processing unit is used for performing variation processing on the intermediate scheduling strategy in the cross strategy population to obtain a variation strategy population.

And the candidate evolution strategy population updating unit is used for determining the variation strategy population as the candidate evolution strategy population obtained by evolution.

And the evolution strategy population determining unit is used for determining the candidate evolution strategy population meeting the evolution conditions as the evolution strategy population and taking the intermediate scheduling strategy in the evolution strategy population as the candidate scheduling strategy.

On the basis of the above technical solution, the mutation processing unit includes:

and the variation probability determining subunit is used for determining the variation probability of each intermediate scheduling strategy in the cross strategy population.

And the state information changing subunit is used for changing the state information in the intermediate dispatching strategy according to a preset rule when the intermediate dispatching strategy is determined to be subjected to variation processing according to the variation probability, wherein the state information represents the response operation of each elevator to the external calling information of each floor.

On the basis of the foregoing technical solution, in an embodiment, the state information changing subunit is specifically configured to determine a first target elevator in an intermediate dispatching strategy when state information of a first target floor in one elevator is changed from elevator calling response information to elevator calling neglect information, where the state information of the first target elevator at the first target floor is elevator calling neglect information; and changing the state information of the first target elevator at the first target floor into elevator calling response information.

On the basis of the above technical solution, in yet another embodiment, the status information changing subunit is specifically configured to read elevator calling marks corresponding to each floor from the external calling information; marking the calling elevator as an empty floor as a second target floor; and changing the state information of each elevator in the intermediate dispatching strategy at the second target floor into neglecting elevator calling.

On the basis of the technical scheme, the evolution condition at least comprises one of the following conditions:

the evolutionary strategy population generating module 540 further comprises:

and the evolution condition judging unit is used for continuously executing the step of selecting the intermediate scheduling strategy of the initial strategy population to obtain a selection strategy population when the evolution condition is not met.

On the basis of the above technical solution, the fitness determining module 550 includes:

the elevator parameter information acquisition unit is used for acquiring elevator parameter information of the elevator;

and the parameter calculation unit is used for calculating the elevator energy consumption value, the average elevator waiting time and the average elevator taking time corresponding to each candidate dispatching strategy according to the elevator parameter information.

And the weighted sum calculating unit is used for calculating the weighted sum of the energy consumption value, the average elevator waiting time and the average elevator taking time corresponding to each candidate scheduling strategy.

And the fitness calculating unit is used for taking the reciprocal of the weighted sum as the fitness of each candidate scheduling strategy.

On the basis of the technical scheme, the elevator parameter information acquisition unit comprises:

and the parameter record searching subunit is used for searching the parameter record corresponding to the serial number of the elevator in a database, and the weighing range of the parameter record is within the preset range of the current weighing of the elevator.

And the average value operator unit is used for calculating the average value of each parameter in the parameter record.

An elevator parameter information combining subunit, configured to combine the average values of the parameters into elevator parameter information of the elevator.

On the basis of the above technical solution, the target scheduling policy determining module 560 includes:

and the sorting unit is used for sorting the candidate scheduling strategies and the initial scheduling strategies from large to small according to the fitness.

And the target scheduling policy determining unit is used for ranking the first scheduling policy in the ranking result as a target scheduling policy.

The device also comprises:

and the dispatching module is used for dispatching the elevator by using the target dispatching strategy after the candidate evolution strategy with the fitness meeting the preset condition is taken as the target dispatching strategy so as to respond to the calling-out information.

On the basis of the technical scheme, the device further comprises:

and the model updating module is used for updating the strategy prediction model by using the target scheduling strategy after the candidate evolution strategy with the fitness meeting the preset condition is taken as the target scheduling strategy.

The product can execute the method provided by any embodiment of the invention, and has corresponding functional modules and beneficial effects of the execution method.

EXAMPLE six

Fig. 6 is a schematic structural diagram of an elevator dispatching strategy processing device according to a sixth embodiment of the present invention. As shown in fig. 6, the elevator dispatching policy processing device includes: a processor 60, a memory 61, an input device 62, and an output device 63. The number of the processors 60 in the elevator dispatching strategy processing device can be one or more, and one processor 60 is taken as an example in fig. 6. The number of the memories 61 in the elevator dispatching strategy processing device can be one or more, and one memory 61 is taken as an example in fig. 6. The processor 60, the memory 61, the input device 62, and the output device 63 of the elevator dispatching strategy processing device can be connected by a bus or other means, and the connection by the bus is exemplified in fig. 6. The elevator dispatching strategy processing equipment can be a computer, a server and the like.

The memory 61 is used as a computer readable storage medium for storing software programs, computer executable programs, and modules, such as program instructions/modules corresponding to the elevator dispatching strategy processing method according to any embodiment of the present invention (e.g., the information obtaining module 510, the initial dispatching strategy determining module 520, the initial strategy population constructing module 530, the evolution strategy population generating module 540, the fitness determining module 550, and the target dispatching strategy determining module 560 in the elevator dispatching strategy processing device). The memory 61 may mainly include a program storage area and a data storage area, wherein the program storage area may store an operating system, an application program required for at least one function; the storage data area may store data created according to use of the device, and the like. Further, the memory 61 may include high speed random access memory, and may also include non-volatile memory, such as at least one magnetic disk storage device, flash memory device, or other non-volatile solid state storage device. In some examples, the memory 61 may further include memory located remotely from the processor 60, which may be connected to the device over a network. Examples of such networks include, but are not limited to, the internet, intranets, local area networks, mobile communication networks, and combinations thereof.

The input device 62 may be used to receive input numeric or character information and generate key signal inputs related to audience user settings and function controls of the elevator dispatch strategy processing apparatus, as well as a camera for capturing images and a sound pickup device for capturing audio data. The output device 63 may include an audio device such as a speaker. It should be noted that the specific composition of the input device 62 and the output device 63 can be set according to actual situations.

The processor 60 executes various functional applications of the device and data processing, i.e., the elevator dispatching strategy processing method described above, by running software programs, instructions and modules stored in the memory 61.

EXAMPLE seven

An embodiment of the present invention further provides a storage medium containing computer-executable instructions, which when executed by a computer processor, are configured to perform a method for elevator dispatching strategy processing, including:

Of course, the storage medium provided by the embodiment of the present invention contains computer-executable instructions, and the computer-executable instructions are not limited to the operation of the elevator dispatching strategy processing method described above, and can also perform related operations in the elevator dispatching strategy processing method provided by any embodiment of the present invention, and have corresponding functions and advantages.

From the above description of the embodiments, it is obvious for those skilled in the art that the present invention can be implemented by software and necessary general hardware, and certainly, can also be implemented by hardware, but the former is a better embodiment in many cases. Based on such understanding, the technical solutions of the present invention may be embodied in the form of a software product, where the computer software product may be stored in a computer-readable storage medium, such as a floppy disk, a Read-only memory (ROM), a Random Access Memory (RAM), a FLASH memory (FLASH), a hard disk or an optical disk of a computer, and includes several instructions to enable a computer device (which may be a robot, a personal computer, a server, or a network device) to execute the elevator scheduling policy processing method according to any embodiment of the present invention.

It should be noted that, in the elevator dispatching strategy processing device, each unit and each module included in the elevator dispatching strategy processing device are only divided according to functional logic, but are not limited to the above division as long as the corresponding functions can be realized; in addition, specific names of the functional units are only for convenience of distinguishing from each other, and are not used for limiting the protection scope of the present invention.

It should be understood that portions of the present invention may be implemented in hardware, software, firmware, or a combination thereof. In the above embodiments, the various steps or methods may be implemented in software or firmware stored in memory and executed by a suitable instruction execution system. For example, if implemented in hardware, as in another embodiment, any one or combination of the following techniques, which are known in the art, may be used: a discrete logic circuit having a logic gate circuit for implementing a logic function on a data signal, an application specific integrated circuit having an appropriate combinational logic gate circuit, a Programmable Gate Array (PGA), a Field Programmable Gate Array (FPGA), or the like.

In the description herein, reference to the term "exemplary," "in an embodiment," or the like, means that a particular feature, structure, material, or characteristic described in connection with the embodiment or example is included in at least one embodiment or example of the invention. In this specification, the schematic representations of the terms used above do not necessarily refer to the same embodiment or example. Furthermore, the particular features, structures, materials, or characteristics described may be combined in any suitable manner in any one or more embodiments or examples.

It is to be noted that the foregoing is only illustrative of the preferred embodiments of the present invention and the technical principles employed. It will be understood by those skilled in the art that the present invention is not limited to the particular embodiments described herein, but is capable of various obvious changes, rearrangements and substitutions as will now become apparent to those skilled in the art without departing from the scope of the invention. Therefore, although the present invention has been described in greater detail by the above embodiments, the present invention is not limited to the above embodiments, and may include other equivalent embodiments without departing from the spirit of the present invention, and the scope of the present invention is determined by the scope of the appended claims.

Claims

1. An elevator dispatching strategy processing method is characterized by comprising the following steps:

carrying out evolution processing on the initial strategy population to obtain an evolution strategy population, and carrying out evolution processing on the initial strategy population to obtain an evolution strategy population, wherein the method comprises the following steps:

determining a candidate evolution strategy population meeting the evolution condition as an evolution strategy population, and taking an intermediate scheduling strategy in the evolution strategy population as a candidate scheduling strategy;

the step of performing variation processing on the intermediate scheduling strategy in the cross strategy population to obtain a variation strategy population comprises the following steps:

when the intermediate scheduling strategy is determined to be subjected to mutation processing according to the mutation probability, state information in the intermediate scheduling strategy is changed according to a preset rule, wherein the state information represents response operation of each elevator to external calling information of each floor;

2. The elevator dispatching strategy processing method of claim 1, wherein the operational state information comprises: the number, the floor, the running direction, the weighing and the internal calling information of the elevator are obtained; the external calling information comprises uplink external calling information and downlink external calling information corresponding to each floor.

3. The elevator dispatching strategy processing method according to claim 1, wherein before inputting the operation state information and the calling-out information into a preset strategy prediction model for strategy analysis to obtain an elevator initial dispatching strategy, the method further comprises:

4. The elevator dispatching strategy processing method according to claim 3, wherein the step of inputting the running state information and the calling-out information into a preset strategy prediction model for strategy analysis to obtain an elevator initial dispatching strategy comprises the following steps:

5. The elevator dispatching strategy processing method of claim 1, wherein constructing an initial strategy population based on the initial dispatching strategy comprises:

6. The elevator dispatching strategy processing method of claim 1, wherein the modifying the state information in the intermediate dispatching strategy according to the preset rule comprises:

7. The elevator dispatching strategy processing method according to claim 1, wherein the changing of the state information in the intermediate dispatching strategy according to the preset rule comprises:

marking the calling elevator as an empty floor as a second target floor;

8. The elevator dispatching strategy processing method of claim 1, wherein the evolving condition comprises at least one of:

9. The elevator dispatching strategy processing method according to claim 1, wherein the determining the fitness of each candidate evolution strategy in the evolution strategy population according to the obtained elevator parameter information comprises:

obtaining elevator parameter information of the elevator;

10. The elevator dispatching strategy processing method of claim 9, wherein the obtaining elevator parameter information for the elevator comprises:

calculating the average value of each parameter in the parameter record;

11. The elevator dispatching strategy processing method according to claim 1, wherein the step of using the candidate evolution strategy with fitness meeting the preset condition as the target dispatching strategy comprises the following steps:

12. The elevator dispatching strategy processing method according to claim 1, characterized in that after the candidate evolution strategy with fitness meeting the preset condition is taken as the target dispatching strategy, the method further comprises the following steps:

updating the policy prediction model using the target scheduling policy.

13. An elevator dispatching strategy processing device is characterized by comprising:

the construction module of the strategy population comprises:

the candidate evolution strategy population initializing unit is used for determining the initial strategy population as an initialized candidate evolution strategy population;

the selection processing unit is used for carrying out selection processing on the intermediate scheduling strategy of the candidate evolution strategy population to obtain a selection strategy population;

the cross processing unit is used for carrying out cross processing on the intermediate scheduling strategies in the selection strategy population to obtain a cross strategy population;

a variation processing unit, configured to perform variation processing on the intermediate scheduling policy in the cross policy population to obtain a variation policy population;

the mutation processing unit includes: a variation probability determination subunit;

the variation probability determining subunit is configured to determine a variation probability of each intermediate scheduling policy in the cross policy population; the state information changing subunit is used for changing the state information in the intermediate dispatching strategy according to a preset rule when the intermediate dispatching strategy is determined to be subjected to mutation processing according to the mutation probability, wherein the state information represents the response operation of each elevator to the external calling information of each floor;

the candidate evolution strategy population updating unit is used for determining the variation strategy population as the candidate evolution strategy population obtained by evolution;

14. An elevator dispatching strategy processing device is characterized by comprising: a memory and one or more processors;

the memory for storing one or more programs;

when executed by the one or more processors, cause the one or more processors to implement the elevator dispatch policy processing method of any of claims 1-12.

15. A storage medium containing computer-executable instructions, which when executed by a computer processor, operate to perform the elevator dispatch policy processing method of any of claims 1-12.