An efficient and low-carbon intelligent warehouse scheduling method

Abstract

The present invention discloses an efficient and low-carbon intelligent warehouse scheduling method, comprising the following steps: pre-setting efficient warehousing and outbound operation areas in the warehouse area; initializing the population size of artificial fish swarms and the maximum number of iterations; encoding each cargo position in the efficient warehousing operation area and the efficient outbound operation area, so that the state of each artificial fish corresponds to the cargo position one by one, and each artificial fish represents a cargo position allocation result; according to the improved artificial fish swarm algorithm, the travel position of the fork equipment in the efficient warehousing operation area and the efficient outbound operation area is planned according to the cargo position allocation result. The present invention pre-arranges the area to be shipped out or to be shipped in, reduces the actual time of warehousing and shipping, and reduces the time occupied by the terminal for the assembly line; effectively utilizes the price difference of wind and valley electricity to achieve an overall energy consumption reduction, thereby achieving the effect of carbon reduction and peak regulation; through the improved artificial fish swarm algorithm, it can effectively avoid falling into the local optimum, improve the real-time performance of the algorithm, and further reduce energy consumption.

Description

Technical Field

[0001] The present invention belongs to the technical field of logistics, and in particular relates to an efficient and low-carbon intelligent warehouse scheduling method. Background Art

[0002] In an automated warehouse that uses fully automatic storage and retrieval equipment (stackers, stacker trucks, forks, etc.), generally speaking, the storage location is fixed once the goods are stored, and the selection of the storage location is generally based on simple principles, such as first-in-first-out strategies. This will actually cause inefficiencies, as fully automatic storage and retrieval equipment always works in the longest travel area, leading to inefficient warehousing and retrieval, and increasing the terminal's occupancy of external shared assembly lines.

[0003] In addition, since the scheduling of fully automatic access equipment often involves multi-objective tasks, swarm intelligence algorithms need to be used for optimization. The artificial fish swarm algorithm in the swarm intelligence algorithm has excellent performance such as simple structure, little influence by initial values, and easy to jump out of local optimality during the optimization process. It is widely used in various problems. Based on this, experts and scholars at home and abroad have successively carried out research on artificial fish swarm algorithms. The step size of traditional artificial fish swarms is fixed, and the convergence speed and accuracy of the algorithm are greatly affected by the step size. A larger step size can effectively improve the convergence speed, but it will reduce the convergence accuracy and easily fall into the local optimal solution. A smaller step size has better convergence accuracy, but the speed is too slow in the early convergence process. The traditional adaptive strategy sets the step size of the artificial fish swarm based on the number of algorithm cycles and the step size factor, ignoring the convergence instability of the artificial fish swarm algorithm, which easily leads to the step size being reduced to a value that does not match the current algorithm before the early algorithm has converged significantly. Summary of the invention

[0004] In view of this, the present application adopts an algorithm to determine the efficient storage area and storage area where the fully automatic storage and retrieval equipment works, adopts the full power setting for in and out storage, and the low power operation setting for storage position conversion, performs equipment storage position conversion when there is no in and out storage task or low electricity price period in the storage area, vacates the storage position to be stored in advance, and moves the equipment to be out of the warehouse to the efficient storage area in advance, thereby improving the efficiency of in and out storage, reducing the waiting time for in and out storage, and saving overall energy consumption to a certain extent. The present application also uses an improved artificial fish school algorithm to find the optimal solution for the scheduling of fully automatic storage and retrieval equipment, further reducing the energy consumption of the equipment.

[0005] To achieve the above objectives, the present application discloses an efficient and low-carbon intelligent warehouse scheduling method, comprising the following steps:

[0006] Pre-set efficient storage and outbound operation areas in the warehouse area, store the equipment to be outbound in the efficient outbound operation area in advance, and when outbound, the rated power consumption of the fork equipment works in the efficient outbound operation area; clear the efficient storage operation area in advance before storage, and when in storage, the rated power consumption of the fork equipment works in the efficient storage operation area;

[0007] Initialize the population size of the artificial fish swarm and the maximum number of iterations;

[0008] Encode each cargo location in the efficient warehouse entry and warehouse exit operation areas, so that the status of each artificial fish corresponds to the cargo location one by one, and each artificial fish represents a cargo location allocation result;

[0009] According to the improved artificial fish swarm algorithm, the travel position of the fork equipment in the efficient warehousing operation area and the efficient outbound operation area is planned according to the cargo location allocation results.

[0010] Furthermore, during the off-peak hours in the smart warehouse or during the off-peak hours of electricity prices, low power consumption is used to sort goods, and full power is used during warehousing and out-of-warehouse operations, thereby reducing the energy consumption of the smart warehouse.

[0011] Furthermore, the area where the fully automatic storage and retrieval equipment is relatively short from the loading and unloading terminal includes at least two directions of travel: the depth of the stacking equipment and the vertical cargo area that can be reached most quickly.

[0012] Furthermore, the goal of the fork scheduling problem is to minimize the total distance and total time of n forks entering and leaving the efficient storage area and the efficient outbound operation area:

[0013]

[0014]

[0015] Among them, X ij is the distance of the i-th fork during the j-th dispatch, T ij is the time consumed by the jth dispatch of the i-th fork, n is the total number of forks, and m is the number of dispatches of a single fork per day;

[0016] The constraints are:

[0017]

[0018]

[0019]

[0020]

[0021] Among them, jis the number of goods entering the efficient storage area during the jth dispatch, is the quantity of goods transported into the efficient storage operation area during the jth dispatch of the i-th fork, Out j is the number of goods entering the outbound efficient operation area during the jth dispatch, Quantity is the quantity of goods transported into the outbound efficient operation area during the jth dispatch of the i-th fork; daily It is the total amount of goods entering and leaving the smart warehouse in one day. When the frequency of a certain product entering the efficient storage operation area or the efficient outbound operation area is higher than the threshold, its position Position is set at the entrance of the efficient storage operation area or the efficient outbound operation area.

[0022] Furthermore, the improved artificial fish school algorithm sets the search range to include a first field of view and a second field of view at the initial stage of the search. The first field of view is a rectangular area of ​​a preset size, and the second field of view is the intersection or approximate intersection of the golden spiral and the dot matrix centered on the position of the artificial fish, thereby performing optimization in a larger range.

[0023] Furthermore, in the middle of the search, after each iteration, the improved artificial fish swarm algorithm clusters all artificial fish using the DBSCAN algorithm to obtain multiple cluster centers and multiple fish swarms, and the step length of the artificial fish in each cluster is:

[0024]

[0025] in is the position of the ith artificial fish, is the location of the cluster center, rand is the random number generation function;

[0026] The field of view of the artificial fish in each cluster is:

[0027]

[0028] n c is the number of artificial fish in the cth cluster fish group.

[0029] Furthermore, a comprehensive evaluation value P is calculated, and the fish schools with high comprehensive evaluation value P are merged with the nearest fish schools;

[0030]

[0031] d ij represents the Euclidean distance between the ith artificial fish and the jth artificial fish in the cth fish cluster, and the comprehensive evaluation value P represents the quotient of the average fitness value of the cth fish cluster and the average Euclidean distance between the artificial fish in the cth fish cluster and other artificial fish.

[0032] Furthermore, in the middle of the search of the artificial fish swarm algorithm, the fish swarm with the largest and second largest comprehensive evaluation value P are cross-mutated respectively. The three artificial fish with the largest fitness in this iteration of each fish swarm are cross-mutated with the three artificial fish with the largest fitness in the closest fish swarm in the genetic algorithm, and the three artificial fish with the worst fitness in the fish swarm are eliminated at the same time.

[0033] In each round of iteration, the fish swarm with the largest and second largest comprehensive evaluation value P are merged with the other nearest fish swarms respectively, and the comprehensive evaluation value P of each fish swarm is calculated in each round of iteration. After multiple iterations, the final two fish swarms are formed, and Gaussian mutation is performed on these two fish swarms. Finally, the position of the fish with the highest fitness in these two fish swarms is the global optimal position, and the other fish swarm moves to this position.

[0034] The beneficial effects of the present invention are as follows:

[0035] 1. The present invention pre-arranges the area to be shipped out or to be shipped in, which reduces the actual shipping time and reduces the time the terminal occupies the assembly line;

[0036] 2. The present invention can effectively utilize the price difference between wind and valley electricity to reduce overall energy consumption, thereby achieving the effect of carbon reduction and peak regulation;

[0037] 3. The present invention can effectively avoid falling into local optimum through the improved artificial fish swarm algorithm, improve the real-time performance of the algorithm, and further reduce energy consumption. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] Figure 1 Flowchart of the present invention;

[0039] Figure 2 The golden spiral image in the embodiment of the present application. DETAILED DESCRIPTION

[0040] The present invention is further described below in conjunction with the accompanying drawings, but the present invention is not limited in any way. Any changes or substitutions made based on the teachings of the present invention belong to the protection scope of the present invention.

[0041] Before beginning to introduce the embodiments of the present application, some terms are introduced.

[0042] 1. Foraging behavior of artificial fish schools: Fish schools choose to move from waters with low food concentration to waters with high food concentration within the field of vision. The mathematical expression of foraging behavior is as follows:

[0043] X j =X i +rand()*visual (1)

[0044]

[0045] Among them, X i is the location where the artificial fish is, visual is the visual field range of the artificial fish, X iprey is the position after the foraging behavior is completed, step is the step length of the artificial fish, and rand() is a random number from 0 to 1. The artificial fish X i If it appears within its visual field range, it will randomly select a state X according to Equation (1) j , and calculate the corresponding function values of them, and compare the calculated function values. If Y j >Y i , then X i will move in the direction of X according to Equation (2) j . Otherwise, the artificial fish X i will randomly select a new state within the visual field and repeat such operations. If the condition is not met yet, the artificial fish will move randomly.

[0046] 2. Aggregation behavior of the artificial fish school: The fish school will gather together, swim from the place with low fitness to the place with high fitness, and the aggregation position should not be too crowded:

[0047]

[0048]

[0049] X in the above formula c represents the central position of the companion fish in the nearby area, n f represents the number of companion fish included in the nearby area, Y i represents the fitness value corresponding to X i , Y c represents the fitness value corresponding to X c , X iswarm represents the new position after the aggregation behavior is executed. When the artificial fish X i is within the visual field range, that is, d ij < visual, search for the number of companion fish in the neighborhood, and then use Equation (3) to obtain the central position X of the companion fish in the neighborhood c . When the minimum value needs to be obtained, Y c / n f <δ (conversely, when the maximum value needs to be obtained, Y c / n f >δ), this indicates that the position state where Xc is located is relatively excellent, the food is abundant and the fish school is not crowded. Therefore, the artificial fish X i will move towards the position where X c is located according to Equation (4). Otherwise, the foraging behavior will be executed.

[0050] 3. Tail-chasing behavior: When a school of fish searches for food, it always swims towards the individual fish with the highest fitness within its field of vision:

[0051]

[0052] X in the above formula m It represents the individual fish with the highest fitness found by the artificial fish in the field of view, X ifollow It is represented by the position after the tail-chasing behavior is completed. i In the field of vision (d ij <visual) Search for the individual X with the highest fitness m , if Y m / n f <δY i If m The state is better, then X i According to formula (5), m The position is moved. Otherwise X i Will engage in foraging behavior.

[0053] The following is an introduction to the efficient and low-carbon intelligent warehouse scheduling method disclosed in the present application. In one embodiment, an efficient warehousing and outbound operation area is preset in the intelligent warehouse, that is, an area where the fully automatic storage and retrieval equipment is a short distance from the warehousing and outbound dock, including at least two directions of travel: the depth of the stacking equipment and the vertical area where the cargo space can be reached most quickly.

[0054] The equipment to be shipped out is stored in advance in the efficient shipping operation area determined by the present invention. When shipping out, the rated power consumption of the fork equipment works in the efficient operation area, thereby improving the shipping efficiency.

[0055] Before warehousing, the efficient warehousing operation area determined by the present invention is cleared in advance. When warehousing, the rated power consumption of the fork equipment works in the efficient operation area, thereby improving the warehousing efficiency.

[0056] In one embodiment, the present application sets the length of the efficient warehousing operation area and the efficient warehousing operation area according to the fork plane stroke, fork plane speed, and fork plane acceleration, that is, the fork plane stroke, fork plane speed, and fork plane acceleration are limited by the length of the efficient warehousing operation area and the efficient warehousing operation area. Set the height of the efficient warehousing operation area and the efficient warehousing operation area according to the fork three-dimensional speed, fork three-dimensional stroke, and fork three-dimensional acceleration, that is, the fork three-dimensional speed, fork three-dimensional stroke, and fork three-dimensional acceleration are limited by the height of the efficient warehousing operation area and the efficient warehousing operation area.

[0057] The goal of the fork scheduling problem is to minimize the total distance and total time of n forks entering and leaving the efficient storage operation area and the efficient storage operation area:

[0058]

[0059]

[0060] Among them, X ij is the distance of the i-th fork during the j-th dispatch, T ij is the time consumed by the jth dispatch of the ith fork, n is the total number of forks, and m is the number of times a single fork is dispatched per day.

[0061] The constraints are:

[0062]

[0063]

[0064]

[0065]

[0066] Among them, j is the number of goods entering the efficient storage area during the jth dispatch, is the quantity of goods transported into the efficient storage operation area during the jth dispatch of the i-th fork, Out j is the number of goods entering the outbound efficient operation area during the jth dispatch, Quantity is the quantity of goods transported into the outbound efficient operation area during the jth dispatch of the i-th fork; daily It is the total amount of goods entering and leaving the smart warehouse in one day. When the frequency of a certain product entering the efficient storage operation area or the efficient outbound operation area is higher than the threshold, its position Position is set at the entrance of the efficient storage operation area or the efficient outbound operation area.

[0067] In one embodiment, the present application uses low power consumption to sort goods during the smart warehouse's off-peak hours or during time-of-use electricity price periods, and uses full power during warehouse entry and exit operations, thereby reducing the smart warehouse's energy consumption and operating electricity costs.

[0068] Since the fork scheduling task is a multi-objective task, this application uses the artificial fish swarm algorithm in the swarm intelligence algorithm to find the best task allocation solution.

[0069] Each cargo location in the efficient inbound operation area and the efficient outbound operation area is coded so that the status of each artificial fish corresponds to the cargo location one by one, and each artificial fish represents a cargo location allocation result. The cargo location includes the operation area type (efficient inbound operation area or efficient outbound operation area), the row and column number of the container, and the cargo location number in the container.

[0070] The artificial fish algorithm is an intelligent optimization method based on the aggregation, foraging, following and random behaviors of fish schools. The algorithm improves its own state according to the current food concentration, sensory distribution, location of partner fish schools, and bulletin board status of the artificial fish. Through these behavioral choices, the artificial fish will eventually converge to a better extreme value.

[0071] When the algorithm tries the optimal value in the search space, if the operator used is not far from the optimal value, then the algorithm can quickly approach the optimal value, but if the operator used is far from the optimal value, then it will take a long time to converge, or even not converge. Therefore, in the early stage of the search, in one embodiment, the search range set by the present application includes a first field of view and a second field of view. The first field of view is a rectangular area of ​​a preset size, and the second field of view is the intersection or approximate intersection (the distance between the points of the dot matrix and the golden spiral is less than the preset distance) of the golden spiral centered on the position of the artificial fish and the dot matrix (that is, the efficient operation area and the efficient storage area are divided into an M×N dot matrix, and the points of the dot matrix are the locations where forklifts store and retrieve goods), and then jump to a larger range to search for the best result. The image of the golden spiral is as follows: Figure 2 As shown, the mathematical formula is as follows:

[0072]

[0073] a and k are preset constants, is the polar angle and ρ is the polar diameter.

[0074] Mid-term search: After each iteration, all artificial fish are clustered using the DBSCAN algorithm to obtain multiple cluster centers and multiple fish schools. The field of view of the artificial fish in each cluster is:

[0075] Step Length

[0076] in is the position of the ith artificial fish, is the location of the cluster center.

[0077] The field of view of the artificial fish in each cluster is:

[0078]

[0079] n c is the number of artificial fish in the cth cluster fish group.

[0080] After multiple fish schools are clustered, each fish school gathers towards the local optimal point. Finally, a selection is made from multiple local optimal points to find the global optimal point, which can effectively avoid falling into the local optimal point and being unable to jump out.

[0081] Calculate the comprehensive evaluation value P, and merge the fish schools with the highest comprehensive evaluation value P into the nearest fish schools;

[0082]

[0083] d ij It represents the Euclidean distance between the ith artificial fish and the jth artificial fish in the cth fish cluster. The comprehensive evaluation value P represents the quotient of the average fitness value of the cth fish cluster and the average Euclidean distance between the artificial fish in the cth fish cluster and other artificial fish. The higher the average fitness value, the closer the fish cluster is to the local optimal value. The smaller the average Euclidean distance between the artificial fish and other artificial fish, the closer the fish in the fish cluster are. Therefore, the larger the comprehensive evaluation value P is, the closer the fish cluster as a whole is to the local optimal value. However, in the middle of the search, multiple fish groups approach multiple local optima respectively. At this time, PK is required to find the global optimum from multiple local optima. Therefore, in one embodiment, the present application further crosses and mutates the fish groups with the largest and second largest comprehensive evaluation values ​​P respectively. The three artificial fish with the largest fitness in this iteration of each fish group and the three artificial fish with the largest fitness in the closest fish group are cross-mutated in the genetic algorithm, and the three artificial fish with the worst fitness in the fish group are eliminated at the same time. In this way, the optimal artificial fish of the nearby fish groups can be learned, and in each round of iteration, the two fish groups are merged with the nearest other fish groups respectively, and the comprehensive evaluation value P of each fish group is calculated in each round of iteration. After multiple iterations, the last two fish groups are formed, and Gaussian mutation is performed on these two. Gaussian mutation is to add a random disturbance term that obeys Gaussian distribution on the basis of the original individual value, which helps individuals jump out of the local optimal solution, thereby enhancing the ability to search for the global optimal solution. Finally, the position of the fish with the highest fitness in the two fish groups is the global optimal, and the other fish group moves to this position, and all artificial fish reach the global optimal.

[0084] The beneficial effects of the present invention are as follows:

[0085] 1. The present invention pre-arranges the area to be shipped out or to be shipped in, which reduces the actual shipping time and reduces the time the terminal occupies the assembly line;

[0086] 2. The present invention can effectively utilize the price difference between wind and valley electricity to reduce overall energy consumption, thereby achieving the effect of carbon reduction and peak regulation;

[0087] 3. The present invention can effectively avoid falling into local optimum through the improved artificial fish swarm algorithm, improve the real-time performance of the algorithm, and further reduce energy consumption.

[0088] As used herein, the word "preferred" is intended to be used as an example, instance, or illustration. Any aspect or design described herein as "preferred" is not necessarily to be construed as being more advantageous than other aspects or designs. On the contrary, the use of the word "preferred" is intended to present concepts in a specific way. The term "or" as used in this application is intended to mean an inclusive "or" rather than an exclusive "or". That is, unless otherwise specified or clear from the context, "X uses A or B" means any one of the naturally included permutations. That is, if X uses A; X uses B; or X uses both A and B, then "X uses A or B" is satisfied in any of the foregoing examples.

[0089] Moreover, although the present disclosure has been shown and described with respect to one or implementations, those skilled in the art will think of equivalent variations and modifications based on the reading and understanding of this specification and the accompanying drawings. The present disclosure includes all such modifications and variations, and is limited only by the scope of the appended claims. In particular, with respect to the various functions performed by the above-mentioned components (such as elements, etc.), the terms used to describe such components are intended to correspond to any component (unless otherwise indicated) that performs the specified function of the component (such as it is functionally equivalent), even if the structure is not equivalent to the disclosed structure of the function in the exemplary implementation of the present disclosure shown herein. In addition, although the specific features of the present disclosure have been disclosed with respect to only one of several implementations, such features can be combined with one or other features of other implementations that may be desired and advantageous for a given or specific application. Moreover, insofar as the terms "including", "having", "containing" or their variations are used in specific embodiments or claims, such terms are intended to be included in a manner similar to the term "comprising".

[0090] The functional units in the embodiments of the present invention may be integrated into a processing module, or each unit may exist physically separately, or multiple or more units may be integrated into one module. The above-mentioned integrated module may be implemented in the form of hardware or in the form of a software functional module. If the integrated module is implemented in the form of a software functional module and sold or used as an independent product, it may also be stored in a computer-readable storage medium. The above-mentioned storage medium may be a read-only memory, a disk or an optical disk, etc. The above-mentioned devices or systems may execute the storage method in the corresponding method embodiment.

[0091] To sum up, the above embodiment is an implementation mode of the present invention, but the implementation mode of the present invention is not limited by the embodiment. Any other changes, modifications, substitutions, combinations, and simplifications that deviate from the spirit and principles of the present invention should be equivalent replacement methods and are included in the protection scope of the present invention.

Claims

1. An efficient and low-carbon intelligent warehouse scheduling method, characterized in that: The following steps are involved: Pre-set efficient storage and outbound operation areas in the warehouse area, and store the equipment to be shipped out in the efficient outbound operation area in advance. When shipping out, the rated power consumption of the fork equipment works in the efficient outbound operation area; Clear the efficient storage operation area in advance before storage. When storage, the rated power consumption of the fork equipment should be in the efficient storage operation area. Initialize the population size of the artificial fish swarm and the maximum number of iterations; Encode each cargo location in the efficient warehouse entry and warehouse exit operation areas, so that the status of each artificial fish corresponds to the cargo location one by one, and each artificial fish represents a cargo location allocation result; According to the improved artificial fish swarm algorithm, the travel position of the fork equipment in the efficient operation area for inbound storage and the efficient operation area for outbound storage is planned according to the cargo space allocation results; In the middle of the search, after each iteration, the improved artificial fish swarm algorithm clusters all artificial fish using the DBSCAN algorithm to obtain multiple cluster centers and multiple fish swarms. The step size of the artificial fish in each cluster is: in It is i Artificial fish in t The location at the moment, is the location of the cluster center, rand is a random number generation function; The field of view of the artificial fish in each cluster is: is the number of artificial fish in the cth cluster of fish.

2. The efficient and low-carbon intelligent warehouse scheduling method according to claim 1 is characterized in that: During the off-peak hours in the smart warehouse or during the off-peak hours of electricity prices, low power consumption is used to sort goods, and full power is used during warehousing and out-of-warehouse operations, thereby reducing the energy consumption of the smart warehouse.

3. The efficient and low-carbon intelligent warehouse scheduling method according to claim 1 is characterized in that: The area where the fully automatic storage and retrieval equipment is relatively short from the loading and unloading dock includes at least two directions of travel: the depth of the stacking equipment and the vertical area where the cargo space can be reached most quickly.

4. The efficient and low-carbon intelligent warehouse scheduling method according to claim 1 is characterized in that: The goal of the fork scheduling problem is n The total distance and total time of each fork entering and exiting the efficient operation area for inbound storage and outbound storage are the shortest: in, X ij It is i The distance of a fork during the jth dispatch, T ij It is i The first j The time consumed by the next scheduling, n is the total number of forks, m is the number of times a single fork is dispatched in a day; The constraints are: in, It is j The number of goods entering the efficient storage operation area during the first dispatch. It is i The first j The number of goods transported into the efficient storage operation area during the next dispatch, It is j The number of goods entering the outbound efficient operation area during the first dispatch. It is i The first j The number of goods transported into the outbound efficient operation area during the dispatch; It is the total amount of goods entering and leaving the smart warehouse in one day; when a certain product enters the efficient operation area for inbound storage or outbound storage, the frequency of the goods entering the efficient operation area for outbound storage is Above threshold threshold When the position Position is set at the entrance of the efficient operation area for incoming storage or the efficient operation area for outgoing storage.

5. The efficient and low-carbon intelligent warehouse scheduling method according to claim 1 is characterized in that: The improved artificial fish swarm algorithm sets the search range to include a first field of view and a second field of view at the initial stage of the search. The first field of view is a rectangular area of ​​a preset size, and the second field of view is the intersection or approximate intersection of the golden spiral and the dot matrix centered at the position of the artificial fish, so as to search for the best result in a larger range.

6. The efficient and low-carbon intelligent warehouse scheduling method according to claim 5 is characterized in that: Calculate the comprehensive evaluation value P, and merge the fish schools with the highest comprehensive evaluation value P into the nearest fish schools; Indicates c The first i Artificial fish and j The Euclidean distance of the artificial fish, the comprehensive evaluation value P represents the c The average fitness of the fish clusters is c The quotient of the average Euclidean distance between the artificial fish of a fish cluster and other artificial fish.

7. The efficient and low-carbon intelligent warehouse scheduling method according to claim 6 is characterized in that: In the middle of the search of the artificial fish swarm algorithm, the fish swarms with the largest and second largest comprehensive evaluation values ​​P are cross-mutated respectively. The three artificial fish with the largest fitness in this iteration of each fish swarm are cross-mutated with the three artificial fish with the largest fitness in the closest fish swarm in the genetic algorithm, and the three artificial fish with the worst fitness in the fish swarm are eliminated at the same time. In each round of iteration, the fish swarm with the largest and second largest comprehensive evaluation value P are merged with the other nearest fish swarms respectively, and the comprehensive evaluation value P of each fish swarm is calculated in each round of iteration. After multiple iterations, the final two fish swarms are formed, and Gaussian mutation is performed on these two fish swarms. Finally, the position of the fish with the highest fitness in these two fish swarms is the global optimal position, and the other fish swarm moves to this position.

Images