Network routing planning method for mine digital twin system based on migration optimization algorithm

Abstract

The present application relates to a method for underlying network routing planning of a mine digital twin system based on a migration algorithm, including: determining k routing planning tasks, the k routing planning tasks include 1 current task and k-1 historically similar tasks; determining the optimization target of each routing planning task, the variable in the optimization target is the routing planning scheme; the optimization target of the k-1 historically similar tasks is a k-1 routing planning scheme evaluation agent model; the optimization targets corresponding to the k routing planning tasks are solved using a multi-objective optimization algorithm to obtain the optimal routing planning scheme for the current task. The present application finds similar problems in historical configuration records based on the underlying routing characteristics that need to be configured at present, and solves them simultaneously with the current network routing planning problem. During the solution process, the mid-way solution result is migrated to the current planning task according to a certain probability, and the migration information is used to accelerate the algorithm solution, which significantly improves the efficiency and quality of the underlying network routing planning of the mine digital twin system.

Description

Technical Field

[0001] The present application relates to the technical field of routing planning, and in particular, to a network routing planning method for a mine digital twin system based on a migration optimization algorithm. Background Art

[0002] At present, the integration of digital twin systems and the digital construction of mines is becoming increasingly close, and the use of digital twin technology to solve mine management and early warning analysis problems has become a trend. The digital twin system is a real-time dynamic hyper-realistic mapping of the physical entity system. It is necessary to detect and analyze the mine situation based on the data of various devices such as sensors. The quality of the underlying network routing is related to the real-time transmission and update quality and speed of data, which plays a vital role in digital twins. Due to the large number of sensor types in mines and the many network transmission methods, coupled with the complex industrial scenes and many equipment and facilities, the network topology is more complex than other network environments. These have brought great difficulties to the underlying network routing planning of the mine digital twin system, resulting in the existing underlying network routing planning methods of the mine digital twin system being ineffective when faced with large-scale network routing planning problems with many devices and complex network topologies. Summary of the invention

[0003] In order to overcome at least one shortcoming in the prior art, an embodiment of the present application provides a network routing planning method for a mine digital twin system based on a migration optimization algorithm.

[0004] In a first aspect, a method for planning underlying network routing of a mine digital twin system based on a migration algorithm is provided, comprising:

[0005] Step S1, determining k routing planning tasks, where the k routing planning tasks include 1 current task and k-1 historical similar tasks;

[0006] Step S2, determining the optimization target of each routing planning task, the variable in the optimization target is the routing planning solution; the optimization target of k-1 historical similar tasks is k-1 routing planning solution evaluation agent models; the optimization target of the current task is the real routing planning solution evaluation function;

[0007] Step S3, using a multi-objective optimization algorithm to solve the optimization objectives corresponding to the k routing planning tasks, and obtain the optimal routing planning solution for the current task.

[0008] In one embodiment, the optimization objectives corresponding to the k routing planning tasks are solved using a multi-objective optimization algorithm to obtain the optimal routing planning solution for the current task, including:

[0009] Step S31, generating an initial population for each routing planning task, the initial population including a plurality of individuals, each individual being a routing planning solution;

[0010] Step S32, for each current population corresponding to the routing planning task, a sub-population is generated by a crossover mutation operator and added to the current population;

[0011] Step S33, determining whether the current iteration number reaches the set iteration number T from the last migration probability update, if so, executing step S34, otherwise, executing step S35;

[0012] Step S34, for each routing planning task, migrate the individuals in the offspring population corresponding to other routing planning tasks to the offspring population corresponding to the current routing planning task according to the current migration probability, obtain the migrated offspring population, and add it to the current population; execute step S35;

[0013] Step S35, for each routing planning task, an individual selection method is used to select excellent individuals in the current population to obtain the next generation population corresponding to each routing planning task;

[0014] Step S36, for each routing planning task, calculate the optimization target value of each individual in the next generation population, and determine whether the number of times the optimization target value is calculated reaches the evaluation setting number maxFE. If it reaches, the process ends. For the current task, the individual with the best optimization target value in the next generation population is the optimal routing planning solution for the current task. Otherwise, execute step S37;

[0015] Step S37, determine whether the current iteration has undergone an individual migration process, if so, update the migration probability and execute step S32, if not, directly execute step S32.

[0016] In one embodiment, for each routing planning task, the optimization target value of each individual in the next generation population is calculated, including:

[0017] For the optimization objectives of k-1 historically similar tasks, each individual in the next generation population is input into k-1 routing planning scheme evaluation agent models, and the output result is the optimization target value of each individual; for the optimization objective of the current task, each individual in the next generation population is input into the true optimization objective function, and the output result is the optimization target value of each individual.

[0018] In one embodiment, for each current population corresponding to each routing planning task, a sub-population is generated through a crossover mutation operator, including:

[0019] Randomly select 2 individuals in the current population and use the simulated binary crossover method to cross and generate 2 offspring individuals. The process of generating offspring individuals is carried out for Second-rate;

[0020] All offspring individuals and individuals in the current population are subjected to polynomial mutation according to the set mutation probability to obtain the offspring population.

[0021] In one embodiment, updating the migration probability includes:

[0022] For any routing planning task, denoted as the current routing planning task, determine the number of individuals a that migrate from the offspring population corresponding to other routing planning tasks j to the offspring population corresponding to the current routing planning task j , j = 1, 2, ..., k-1, the total number of individuals migrating

[0023] Determine the number of individuals b that are retained in the next generation population after the individual selection method is selected and migrated from the sub-generation population corresponding to other routing planning tasks j ;

[0024] According to the number of individuals j and the number of individuals b j Determine the retention ratio r j ;

[0025] Determine the number of individuals to be migrated d = nn ′ , where n is the total number of migrating individuals;

[0026] According to the number of individuals to be migrated d and r j , determine the update probability:

[0027]

[0028] Among them, P j is the probability of migrating individuals from the offspring population corresponding to the jth routing planning task to the offspring population corresponding to the current routing planning task.

[0029] In the second aspect, a bottom network routing planning device for a mine digital twin system based on a migration algorithm is provided, comprising:

[0030] A routing planning task determination module is used to determine k routing planning tasks, where the k routing planning tasks include 1 current task and k-1 historical similar tasks;

[0031] The optimization target determination module is used to determine the optimization target of each routing planning task. The variable in the optimization target is the routing planning solution. The optimization targets of k-1 historically similar tasks are k-1 routing planning solution evaluation agent models, and the optimization target of the current task is the real routing planning solution evaluation function.

[0032] The optimization target solving module is used to solve the optimization targets corresponding to k routing planning tasks using a multi-objective optimization algorithm to obtain the optimal routing planning solution for the current task.

[0033] In one embodiment, the optimization target solving module is also used to implement the following functions:

[0034] Step S31, generating an initial population for each routing planning task, the initial population including a plurality of individuals, each individual being a routing planning solution;

[0035] Step S32, for each current population corresponding to the routing planning task, a sub-population is generated by a crossover mutation operator and added to the current population;

[0036] Step S33, determining whether the current iteration number reaches the set iteration number T from the last migration probability update, if so, executing step S34, otherwise, executing step S35;

[0037] Step S34, for each routing planning task, migrate the individuals in the offspring population corresponding to other routing planning tasks to the offspring population corresponding to the current routing planning task according to the current migration probability, obtain the migrated offspring population, and add it to the current population;

[0038] Step S35, for each routing planning task, an individual selection method is used to select excellent individuals in the current population to obtain the next generation population corresponding to each routing planning task;

[0039] Step S36, for each routing planning task, calculate the optimization target value of each individual in the next generation population, and determine whether the number of times the optimization target value is calculated reaches the evaluation setting number maxFE. If it reaches, the process ends. For the current task, the individual with the best optimization target value in the next generation population is the optimal routing planning solution for the current task. Otherwise, execute step S37;

[0040] Step S37, determine whether the current iteration has undergone an individual migration process, if so, update the migration probability and execute step S32, if not, directly execute step S32.

[0041] In one embodiment, the optimization target solving module is further used to:

[0042] For the optimization objectives of k-1 historically similar tasks, each individual in the next generation population is input into k-1 routing planning scheme evaluation agent models, and the output result is the optimization target value of each individual; for the optimization objective of the current task, each individual in the next generation population is input into the true optimization objective function, and the output result is the optimization target value of each individual.

[0043] In one embodiment, the optimization target solving module is further used to:

[0044] Randomly select 2 individuals in the current population and use the simulated binary crossover method to cross and generate 2 offspring individuals. The process of generating offspring individuals is carried out for Second-rate;

[0045] All offspring individuals and individuals in the current population are subjected to polynomial mutation according to the set mutation probability to obtain the offspring population.

[0046] In one embodiment, the optimization target solving module is further used to:

[0047] For any routing planning task, denoted as the current routing planning task, determine the number of individuals a that migrate from the offspring population corresponding to other routing planning tasks j to the offspring population corresponding to the current routing planning task j , j = 1, 2, ..., k-1, the total number of individuals migrating

[0048] Determine the number of individuals b that are retained in the next generation population after the individual selection method is selected and migrated from the sub-generation population corresponding to other routing planning tasks j ;

[0049] According to the number of individuals j and the number of individuals b j Determine the retention ratio r j ;

[0050] Determine the number of individuals to be migrated d = nn ′ , where n is the total number of migrating individuals;

[0051] According to the number of individuals to be migrated d and r j , determine the update probability:

[0052]

[0053] Among them, P j is the probability of migrating individuals from the offspring population corresponding to the jth routing planning task to the offspring population corresponding to the current routing planning task.

[0054] Compared with the prior art, this application has the following beneficial effects:

[0055] This application finds similar problems in historical configuration records based on the characteristics of the underlying routing that currently needs to be configured, and solves them simultaneously with the current network routing planning problem. During the solution process, the mid-solution results are migrated to the current planning task according to a certain probability, and the migration information is used to accelerate the algorithm solution, which significantly improves the efficiency and quality of the underlying network routing planning of the mine digital twin system. BRIEF DESCRIPTION OF THE DRAWINGS

[0056] The present application may be better understood by referring to the following description given in conjunction with the accompanying drawings, which together with the following detailed description are included in this specification and form a part of this specification. In the drawings:

[0057] Figure 1 A flowchart of a method for underlying network routing planning of a mine digital twin system based on a migration algorithm according to an embodiment of the present application is shown;

[0058] Figure 2 A structural block diagram of an underlying network routing planning device for a mine digital twin system based on a migration algorithm according to an embodiment of the present application is shown. DETAILED DESCRIPTION

[0059] The exemplary embodiments of the present application will be described below in conjunction with the accompanying drawings. For the sake of clarity and conciseness, not all features of the actual embodiments are described in the specification. However, it should be understood that many implementation-specific decisions can be made in the process of developing any such actual embodiments in order to achieve the specific goals of the developer, and these decisions may vary from embodiment to embodiment.

[0060] It is also necessary to explain here that, in order to avoid obscuring the present application due to unnecessary details, only the device structure closely related to the scheme according to the present application is shown in the drawings, while other details that are not closely related to the present application are omitted.

[0061] It should be understood that the present application is not limited to the described implementation forms due to the following description with reference to the accompanying drawings. In this article, where feasible, the embodiments can be combined with each other, features between different embodiments can be replaced or borrowed, and one or more features can be omitted in one embodiment.

[0062] The present application embodiment provides a method for planning underlying network routing of a mine digital twin system based on a migration algorithm. Figure 1 A flowchart of a method for planning underlying network routing for a mine digital twin system based on a migration algorithm according to an embodiment of the present application is shown. Figure 1 , methods include:

[0063] Step S1, determine k routing planning tasks, the k routing planning tasks include 1 current task and k-1 historical similar tasks; here, the historical tasks and the current task can be evaluated according to the scale of the routing topology, the type of equipment, the number of equipment, etc., and k-1 historical similar tasks that are closest to the evaluation results of the current task are selected from the historical tasks. Specifically, k can be set to 10.

[0064] Step S2, determine the optimization target of each routing planning task. Here, the optimization target can be selected based on network reliability, soft delay requirements, path cost, load balancing, rapid diversion of data packets arriving due to distribution link failure, etc., and set constraints for the optimization target. Here, the constraints can be routing node constraints.

[0065] In this step, the variable in the optimization objective is the routing plan. This application minimizes each task at the same time. The optimization objective corresponding to each task is a multi-objective optimization problem:

[0066]

[0067] Among them, X k represents the variable vector corresponding to the kth task, F1(x1),…,F k (x k ) represents k tasks, is the m1 optimization goal corresponding to the first task, is the m2 optimization goal corresponding to the second task, is the m corresponding to the kth task k An optimization goal.

[0068] Step S3, using a multi-objective optimization algorithm to solve the optimization objectives corresponding to the k routing planning tasks, and obtain the optimal routing planning solution for the current task.

[0069] In this embodiment, the optimization target of k-1 historical similar tasks is k-1 routing plan evaluation proxy models; the optimization target of the current task is the real routing plan evaluation function. The construction method of the routing plan evaluation proxy model is:

[0070] The planning schemes obtained by solving k-1 historical similar tasks are used as the input data of the model, and the optimization target values ​​corresponding to the planning schemes are used as the output data to construct the data set;

[0071] Select a suitable proxy model, input the data set into the proxy model for training, and obtain the trained proxy model, which is the k-1 routing planning scheme evaluation proxy model. Here, the proxy model can be selected from existing classic proxy models such as Kriging, radial basis function network, BP neural network, etc.

[0072] In this embodiment, similar problems in historical configuration records are found based on the characteristics of the underlying routing that currently needs to be configured, and are solved simultaneously with the current network routing planning problem. During the solution process, the mid-solution results are migrated to the current planning task according to a certain probability, and the migration information is used to accelerate the algorithm solution, thereby significantly improving the efficiency and quality of the underlying network routing planning of the mine digital twin system.

[0073] In one embodiment, in step S3, all optimization objectives are solved using a multi-objective optimization algorithm to obtain an optimal routing planning solution for the current task, including:

[0074] Step S31, generating an initial population for each routing planning task, the initial population includes multiple individuals, each individual is a routing planning scheme; here, the construction of the initial population can be generated in a random manner, or can be generated according to the rules of network routing planning.

[0075] Step S32, for the initial population, generate a child population through the crossover mutation operator and add it to the current population; the initial population size is 100.

[0076] Step S33, determining whether the current iteration number reaches the set number T from the last migration probability update, where T can be 10. If so, executing step S34, otherwise, executing step S35;

[0077] Step S34: for each routing planning task, individuals in the offspring population corresponding to other routing planning tasks are migrated to the offspring population corresponding to the current routing planning task according to the current migration probability, and the migrated offspring population is obtained and added to the current population; here, the current migration probability is P j , j = 1, 2, ..., k-1, j is the probability of migrating individuals in the offspring population corresponding to the jth routing planning task to the offspring population corresponding to the current routing planning task; for the current routing planning task, the total number of individuals migrated from all other routing planning tasks is set to n, and n can be 30; specifically, an automatic encoder can be used to migrate individuals, and step S35 is executed; in this step, the current migration probability P j The initial value is The initial value of the number of times the optimization target value is calculated is 0.

[0078] Step S35, for each routing planning task, an individual selection method is used to select excellent individuals in the current population to obtain the next generation population corresponding to each routing planning task. Here, a non-dominated sorting method can be used.

[0079] Step S36, for each routing planning task, calculate the optimization target value of each individual in the next generation population, and determine whether the number of times the optimization target value is calculated reaches the evaluation setting number maxFE, where the evaluation setting number maxFE is 100000. If it reaches, the process ends. For the current task, the individual with the best optimization target value in the next generation population is the optimal routing planning solution for the current task. Here, the individual with the best optimization target value can be the individual with the smallest optimization target value. Otherwise, execute step S37;

[0080] Step S37, determine whether the current iteration has undergone an individual migration process, if so, update the migration probability and execute step S32, if not, directly execute step S32.

[0081] In the above embodiment, when the optimization target of k-1 historically similar tasks is k-1 routing plan evaluation proxy models, in step S36, for each routing plan task, the optimization target value of each individual in the next generation population is calculated. When calculating k-1 historically similar tasks, each individual in the next generation population can be input into the k-1 routing plan evaluation proxy models, and the output result is the optimization target value of each individual. For the optimization target of the current task, each individual in the next generation population is input into the real optimization objective function, and the output result is the optimization target value of each individual.

[0082] In one embodiment, for each current population corresponding to each routing planning task, a sub-population is generated through a crossover mutation operator, including:

[0083] Randomly select 2 individuals in the current population and use the simulated binary crossover method to cross and generate 2 offspring individuals. The process of generating offspring individuals is carried out for times; here the crossover probability can be 0.8,

[0084] All offspring individuals and individuals in the current population are subjected to polynomial mutation according to the set mutation probability to obtain the offspring population. Here, the set mutation probability is 0.6.

[0085] In one embodiment, updating the migration probability includes:

[0086] For any routing planning task, denoted as the current routing planning task, determine the number of individuals a that migrate from the offspring population corresponding to other routing planning tasks j to the offspring population corresponding to the current routing planning task j , j = 1, 2, ..., k-1, the total number of individuals migrating

[0087] Determine the number of individuals b that are retained in the next generation population after the individual selection method is selected and migrated from the sub-generation population corresponding to other routing planning tasks j ;

[0088] According to the number of individuals j and the number of individuals b j Determine the retention ratio r j ;

[0089] Determine the number of individuals to be migrated d = nn ′ , where n is the total number of migrating individuals;

[0090] According to the number of individuals to be migrated d and r j , update the migration probability:

[0091]

[0092] Among them, P j ′ is the updated migration probability of migrating individuals from the offspring population corresponding to the jth routing planning task to the offspring population corresponding to the current routing planning task;

[0093] The invention concept is the same as that of the underlying network routing planning method of the mine digital twin system based on the migration algorithm. This embodiment also provides a corresponding underlying network routing planning device of the mine digital twin system based on the migration algorithm. Figure 2 The structural block diagram of the underlying network routing planning device of the mine digital twin system based on the migration algorithm according to the embodiment of the present application is shown, and the device includes:

[0094] A routing planning task determination module 21 is used to determine k routing planning tasks, where the k routing planning tasks include 1 current task and k-1 historical similar tasks;

[0095] The optimization target determination module 22 is used to determine the optimization target of each routing planning task, and the variable in the optimization target is the routing planning solution; the optimization target of k-1 historical similar tasks is k-1 routing planning solution evaluation agent models; the optimization target of the current task is the real routing planning solution evaluation function;

[0096] The optimization target solving module 23 is used to solve the optimization targets corresponding to the k routing planning tasks using a multi-objective optimization algorithm to obtain the optimal routing planning solution for the current task.

[0097] The underlying network routing planning device for a mine digital twin system based on a migration algorithm in an embodiment of the present application, the specific implementation functions of each module are the same as the underlying network routing planning method for a mine digital twin system based on a migration algorithm in the aforementioned embodiment, and will not be described in detail again.

[0098] The above are only various implementations of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art who is familiar with the present technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

Claims

1. A method for underlying network routing planning of a mine digital twin system based on a migration algorithm, characterized in that: include: Step S1, determining k routing planning tasks, wherein the k routing planning tasks include 1 current task and k-1 historical similar tasks; Step S2, determining the optimization target of each routing planning task, the variable in the optimization target is the routing planning scheme; the optimization target of the k-1 historically similar tasks is the k-1 routing planning scheme evaluation agent model; the optimization target of the current task is the real routing planning scheme evaluation function; Step S3, using a multi-objective optimization algorithm to solve the optimization objectives corresponding to the k routing planning tasks, and obtaining the optimal routing planning solution for the current task; The optimization objectives corresponding to the k routing planning tasks are solved by a multi-objective optimization algorithm to obtain the optimal routing planning solution for the current task, including: Step S31, generating an initial population for each routing planning task, wherein the initial population includes a plurality of individuals, and each individual is a routing planning solution; Step S32, for each current population corresponding to the routing planning task, a sub-population is generated by a crossover mutation operator and added to the current population; Step S33, determining whether the current iteration number reaches the set iteration number T from the last migration probability update, if so, executing step S34, otherwise, executing step S35; Step S34, for each routing planning task, migrate the individuals in the offspring population corresponding to other routing planning tasks to the offspring population corresponding to the current routing planning task according to the current migration probability, obtain the migrated offspring population, and add it to the current population; execute step S35; Step S35, for each routing planning task, an individual selection method is used to select excellent individuals in the current population to obtain the next generation population corresponding to each routing planning task; Step S36, for each routing planning task, calculate the optimization target value of each individual in the next generation population, and determine whether the number of times the optimization target value is calculated reaches the evaluation setting number maxFE. If so, the process ends. For the current task, the individual with the best optimization target value in the next generation population is the optimal routing planning solution for the current task. Otherwise, execute step S37; Step S37, determining whether the current iteration undergoes an individual migration process, if so, updating the migration probability and executing step S32, if not, directly executing step S32.

2. The method according to claim 1, characterized in that in, For each routing planning task, the optimization target value of each individual in the next generation population is calculated, including: For the optimization objectives of k-1 historically similar tasks, each individual in the next generation population is input into k-1 routing planning scheme evaluation agent models, and the output result is the optimization target value of each individual; for the optimization objective of the current task, each individual in the next generation population is input into the true optimization objective function, and the output result is the optimization target value of each individual.

3. The method according to claim 1, characterized in that in, For the current population corresponding to each routing planning task, the offspring population is generated through the crossover mutation operator, including: Randomly select 2 individuals from the current population and use the simulated binary crossover method to cross over to generate 2 offspring individuals. The process of generating offspring individuals is performed for Second-rate; All offspring individuals and individuals in the current population are subjected to polynomial mutation according to a set mutation probability to obtain the offspring population.

4. The method according to claim 1, characterized in that The updating of the migration probability comprises: For any routing planning task, denoted as the current routing planning task, determine the number of individuals a that migrate from the offspring population corresponding to other routing planning tasks j to the offspring population corresponding to the current routing planning task j , j = 1, 2, ..., k-1, the total number of individuals migrating Determine the number of individuals b that are retained in the next generation population after selection by the individual selection method and migrated from the child population corresponding to other routing planning tasks j ; According to the number of individuals a j and the number of individuals b j Determine the retention ratio r j ; Determine the number of individuals to be migrated d = nn ′ , where n is the total number of migrating individuals; According to the number of individuals to be migrated d and r j , determine the update probability: Among them, P j is the probability of migrating individuals from the offspring population corresponding to the jth routing planning task to the offspring population corresponding to the current routing planning task.

5. A bottom network routing planning device for a mine digital twin system based on a migration algorithm, characterized in that: include: A routing planning task determination module is used to determine k routing planning tasks, wherein the k routing planning tasks include 1 current task and k-1 historical similar tasks; An optimization target determination module is used to determine the optimization target of each routing planning task, wherein the variable in the optimization target is the routing planning solution; the optimization targets of the k-1 historically similar tasks are k-1 routing planning solution evaluation proxy models, and the optimization target of the current task is the real routing planning solution evaluation function; An optimization target solving module is used to solve the optimization targets corresponding to the k routing planning tasks using a multi-objective optimization algorithm to obtain the optimal routing planning solution for the current task; The optimization target solving module is also used to realize the following functions: Step S31, generating an initial population for each routing planning task, wherein the initial population includes a plurality of individuals, and each individual is a routing planning solution; Step S32, for each current population corresponding to the routing planning task, a sub-population is generated by a crossover mutation operator and added to the current population; Step S33, determine whether the current iteration number reaches the set iteration number T from the last migration probability update, if so, execute step S34, otherwise, execute step S35; Step S34, for each routing planning task, migrate the individuals in the offspring population corresponding to other routing planning tasks to the offspring population corresponding to the current routing planning task according to the current migration probability, obtain the migrated offspring population, and add it to the current population; Step S35, for each routing planning task, an individual selection method is used to select excellent individuals in the current population to obtain the next generation population corresponding to each routing planning task; Step S36, for each routing planning task, calculate the optimization target value of each individual in the next generation population, and determine whether the number of times the optimization target value is calculated reaches the evaluation setting number maxFE. If so, the process ends. For the current task, the individual with the best optimization target value in the next generation population is the optimal routing planning solution for the current task. Otherwise, execute step S37; Step S37, determining whether the current iteration undergoes an individual migration process, if so, updating the migration probability and executing step S32, if not, directly executing step S32.

6. The device according to claim 5, characterized in that The optimization target solving module is also used for: For the optimization objectives of k-1 historically similar tasks, each individual in the next generation population is input into k-1 routing planning scheme evaluation agent models, and the output result is the optimization target value of each individual; for the optimization objective of the current task, each individual in the next generation population is input into the true optimization objective function, and the output result is the optimization target value of each individual.

7. The device according to claim 5, characterized in that The optimization target solving module is also used for: Randomly select 2 individuals from the current population and use the simulated binary crossover method to cross over to generate 2 offspring individuals. The process of generating offspring individuals is performed for Second-rate; All offspring individuals and individuals in the current population are subjected to polynomial mutation according to a set mutation probability to obtain the offspring population.

8. The device according to claim 5, characterized in that The optimization target solving module is also used for: For any routing planning task, denoted as the current routing planning task, determine the number of individuals a that migrate from the offspring population corresponding to other routing planning tasks j to the offspring population corresponding to the current routing planning task j , j = 1, 2, ..., k-1, the total number of individuals migrating Determine the number of individuals b that are retained in the next generation population after selection by the individual selection method and migrated from the child population corresponding to other routing planning tasks j ; According to the number of individuals a j and the number of individuals b j Determine the retention ratio r j ; Determine the number of individuals to be migrated d = nn ′ , where n is the total number of migrating individuals; According to the number of individuals to be migrated d and r j , determine the update probability: Among them, P j is the probability of migrating individuals from the offspring population corresponding to the jth routing planning task to the offspring population corresponding to the current routing planning task.

Images