Multi-vehicle task allocation and path optimization simulation platform and its implementation method
By designing a multi-vehicle task allocation and path optimization simulation platform, using topological relationship-based genetic algorithms and multi-threaded architectures, the problem of high time complexity in the existing technology is solved, and the online approximation solution and visual simulation of vehicle path problems are realized. , suitable for large-scale node computing, improving the computing efficiency and scalability of the system.
Patent Information
- Application Number
- CN202011400672.3
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2020-12-02
- Publication Date
- 2024-11-26
- Estimated Expiration
- 2040-12-02
AI Technical Summary
When facing large-scale data and complex constraints, the existing multi-vehicle task allocation and path optimization algorithms have high time complexity and cannot effectively meet the actual application needs. The existing simulation platform lacks online approximation solutions and visualization capabilities. It is difficult to realize visualization of algorithm operation process and real-time demonstration of logistics solutions.
A multi-vehicle task allocation and path optimization simulation platform was designed, including the business application layer, technology layer and data layer. Through the order entry system, algorithm matching module and path planning simulation system, it uses genetic algorithms based on topological relationships and multi-threaded architecture. , realize the best algorithm matching for vehicle path problems and real-time demonstration of logistics solutions, reducing the algorithm time complexity.
It significantly reduces the time complexity of the algorithm, provides online approximation solutions and visual simulation capabilities, is suitable for large-scale node computing, improves the computing efficiency and scalability of the system, and can be quickly applied in actual scenarios.
Smart Images

Figure CN114580695B_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of vehicle path planning, and in particular to a multi-vehicle task allocation and path optimization simulation platform and an implementation method thereof. Background Art
[0002] Multi-vehicle task allocation and path optimization are research hotspots in the fields of robotics, logistics and computer applications. The multi-vehicle task allocation and path optimization simulation platform mainly involves the study of vehicle routing problems (VRP) and various constrained problems extended from them, such as the capacity-constrained VRP problem (CVRP), the vehicle routing problem with time windows (VRP-tw), the vehicle routing problem with pick and dispatch (VRP-pd), etc.
[0003] The VRP problem is one of the most classic optimization problems in the field of operations research and optimization. It refers to the situation where m vehicles starting from a warehouse will deliver a certain amount of goods to n customers. When serving a group of users, determine the best route for a group of vehicles to travel. The goal is to minimize the overall transportation cost. The solution to the classic VRP problem is a set of paths that all start from the warehouse, all customers are served only once and then return to the warehouse, and the vehicle load constraints are met.
[0004] Existing research on this issue is based on the VRP problem and the extension of the problem and related algorithms. It is too focused on a single limited problem and cannot provide a reasonable optimization strategy for the practical problems faced by logistics companies that keep pace with the times. Moreover, the limitations of existing algorithms cannot meet the application of actual scenarios well. For example, the ATS-MD algorithm requires topological sorting of all data, and its time complexity is very high. If the amount of data is huge, the ATS-MD algorithm will take a lot of time, which is not conducive to task planning and scheduling; although the DP-TR algorithm is an accurate solution algorithm, it also considers the constraints of topological relationships when using dynamic programming technology, which effectively reduces the time complexity of the algorithm, but the time complexity is still exponential, which is not conducive to the calculation of large-scale nodes. These algorithms are obviously not applicable to VRP-pd problems in actual situations.
[0005] The simulation platform is an effective tool for studying task allocation and path planning. Domestic and foreign research institutions are lacking in this aspect when studying task allocation and path planning. The existing multi-vehicle simulation platform mainly provides a systematic framework for optimization strategies. When targeting specific application areas, it focuses on communication between multiple robots and cannot realize the visualization of the algorithm operation process and the logistics plan distribution process. In the field of logistics warehousing environment, there are few simulation platforms that can provide online approximate solutions for problems such as VRP. They almost do not consider the expansion of constraints such as the ratio of orders and vehicles, time windows, and the scalability of storage scale and vehicle load scale. The modeling application process is complicated, and the time complexity of the algorithm is still large, which is not practical. Summary of the invention
[0006] In view of the shortcomings of the prior art, the present invention provides a multi-vehicle task allocation and path optimization simulation platform and its implementation method, which can provide online approximate solutions for problems such as VRP, realize visual simulation of the algorithm operation process and logistics plan distribution process, and significantly reduce the time complexity of the algorithm.
[0007] In order to achieve the above-mentioned purpose, the present invention adopts the following technical solutions:
[0008] A multi-vehicle task allocation and path optimization simulation platform, comprising:
[0009] Business application layer, including order entry system and route planning simulation system;
[0010] The technical layer includes the algorithm matching module;
[0011] The data layer includes a system configuration database, wherein the system configuration database stores a plurality of system built-in algorithms;
[0012] The order entry system is used to interact with the operator, the algorithm matching module is used to select the best algorithm for the vehicle path problem that is suitable for the operator from the system's built-in algorithms based on the data information input by the operator, and the path planning simulation system is used to demonstrate the logistics plan distribution process in real time based on the best algorithm.
[0013] As one implementation method, the order entry system interacts with the operator, including selecting a vehicle routing problem that needs to be visualized, and a pre-processing step of importing customer point information, order demand information and vehicle warehouse information into the system in real time.
[0014] As one of the implementation modes, the technical layer also includes a research problem expansion module, which is used to introduce constraint parameter conditions of an extended problem based on the vehicle path problem in real time. The algorithm matching module is also used to select the best algorithm for the extended problem of the vehicle path problem that is suitable for the system from the built-in algorithms of the system according to the data information input by the operator.
[0015] As one implementation method, the path planning simulation system is also used to display the data information input by the operator and the operation process of the algorithm in real time.
[0016] As one of the implementation modes, the technical layer also includes a data preprocessing module and a multi-threaded algorithm operation module. The data preprocessing module is used to determine whether the calculation amount of the matching best algorithm process is huge, and the multi-threaded algorithm operation module is used to enable the multi-threaded architecture when the calculation amount of the matching best algorithm process exceeds a threshold.
[0017] As one of the implementation modes, an extension of the vehicle routing problem includes a loading and unloading vehicle routing problem, and the optimal algorithm for the loading and unloading vehicle routing problem is a genetic algorithm based on topological relationships. The genetic algorithm based on topological relationships generates offspring through selection, crossover and mutation, and maintains the constraints of topological relationships during each operation.
[0018] As one implementation mode, the genetic algorithm based on topological relationship includes:
[0019] Initialize the population: randomly select path nodes with zero in-degree and sort them, maintaining the constraints of topological relationships;
[0020] Crossover: Select two path solutions that meet the topological sorting from the population for crossover;
[0021] Mutation: Randomly select a path exchange node, determine the exchange range that meets the conditions, and exchange it with another path node randomly selected within the exchange range to generate a new path plan;
[0022] Repeat the above iterative process until the total distance of the new path solution is the shortest.
[0023] Another object of the present invention is to provide a method for implementing a multi-vehicle task allocation and path optimization simulation platform, comprising:
[0024] Interact with the operator;
[0025] According to the data information input by the operator, the best algorithm for the vehicle routing problem is selected from the system's built-in algorithms;
[0026] The logistics plan distribution process is demonstrated in real time based on the optimal algorithm.
[0027] As one implementation manner, the step of interacting with the operator includes:
[0028] Select the vehicle routing problem that needs to be visualized;
[0029] Import customer point information into the system;
[0030] Import vehicle storage information into the system;
[0031] Import order demand information into the system.
[0032] As one implementation method, the process of selecting the best algorithm further includes:
[0033] It is determined whether the computation amount of the best matching algorithm process is large, and a multi-thread architecture is enabled when the computation amount of the best matching algorithm process exceeds a threshold.
[0034] The multi-vehicle task allocation and path optimization simulation platform of the present invention has three layers: business application layer, technical layer, and data layer. After the operator selects a specific vehicle path problem according to the needs, the system matches the corresponding optimal algorithm according to the problem, and can perform real-time demonstration of the vehicle's logistics plan distribution process according to the optimal algorithm. It solves the shortcomings of existing vehicle task allocation and path optimization application fields, such as lack of pertinence, complex simulation process, poor scalability, and weak purposefulness, significantly reduces the algorithm's time complexity, and can provide a visual verification platform for various algorithms with the shortest distance as the optimization goal. BRIEF DESCRIPTION OF THE DRAWINGS
[0035] Figure 1 A schematic diagram of the architecture of a multi-vehicle task allocation and path optimization simulation platform according to an embodiment of the present invention;
[0036] Figure 2 is an application flow chart of a genetic algorithm based on topological relationships according to an embodiment of the present invention;
[0037] Figure 3 A schematic diagram of a demonstration interface of a simulation system according to an embodiment of the present invention;
[0038] Figure 4 A schematic diagram of a flow chart of a method for implementing a simulation platform according to an embodiment of the present invention;
[0039] Figure 5 The vehicle task simulation execution flow chart of an embodiment of the present invention. DETAILED DESCRIPTION
[0040] In the present invention, the terms "disposed", "provided with", and "connected" should be understood in a broad sense. For example, it can be a fixed connection, a detachable connection, or an integral structure; it can be a mechanical connection, or an electrical connection; it can be a direct connection, or an indirect connection through an intermediate medium, or it can be an internal connection between two devices, elements, or components. For those of ordinary skill in the art, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.
[0041] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.
[0042] See also Figure 1 The embodiment of the present invention provides a multi-vehicle task allocation and path optimization simulation platform, including a business application layer 1, a technical layer 2 and a data layer 3, wherein the business application layer 1 includes an order entry system 11 and a path planning simulation system 12, the technical layer 2 includes an algorithm matching module 21, and the data layer 3 includes a system configuration database 31, and the system configuration database 31 stores a variety of system built-in algorithms.
[0043] The order entry system 11 can interact with the operator, the algorithm matching module 21 can select the best algorithm for the VRP problem suitable for the operator from the system's built-in algorithms based on the data information input by the operator, and the path planning simulation system 12 can demonstrate the logistics plan distribution process in real time based on the best algorithm.
[0044] The data information input by the operator may mainly include VRP problems that need to be visualized, customer point information, vehicle warehouse information, and order demand information. The VRP problems that need to be visualized are simulation problems that the operator needs to visualize, customer point information is customer coordinates, vehicle warehouse information includes vehicle and warehouse coordinates, quantity and load capacity, etc., and order demand information is the cargo distribution plan between users.
[0045] The above-mentioned process of interacting with the operator through the order entry system 11 may specifically include: selecting the VRP problem that needs to be visualized, and a pre-processing step of importing customer point information, order demand information and vehicle storage information into the system in real time.
[0046] The path planning simulation system 12 can have a variety of real-time demonstration contents, which can display the data information input by the operator, the operation process of the algorithm, and the real-time demonstration of the logistics plan distribution process, which can well show the audience the specific path planning methods, processes and characteristics.
[0047] In addition to the algorithm matching module 21, the technical layer 2 can also include a research problem expansion module 22. The research problem expansion module 22 can be used to introduce constraint parameter conditions of extended problems based on the VRP problem in real time, and the algorithm matching module 21 can also be used to select the best algorithm for the extended problem of VRP that is adapted to it from the system built-in algorithm according to the data information input by the operator. The research problem expansion module 22 can provide scalability for the simulation platform, and can be used for real-time research on various extended problems based on the VRP problem. In actual applications, the system built-in algorithms corresponding to various extended problems can be stored in the system configuration database 31, so that the algorithm matching module 21 can adapt to different VRP problems and their extended problems, successfully complete the simulation demonstration process, realize the customization of personalized needs, and expand the application scenarios of the platform.
[0048] In addition, the technical layer 2 may also include a data preprocessing module 23 and a multi-threaded algorithm operation module 24. The data preprocessing module 23 is used to determine whether the calculation amount of the matching best algorithm process is huge, and the multi-threaded algorithm operation module 24 is used to enable the multi-threaded architecture when the calculation amount of the matching best algorithm process exceeds a threshold.
[0049] Among them, the processing result of the data preprocessing module 23 will be beneficial to the operation of the algorithm and the module call of the platform to accelerate the operation process of the simulation platform. For example, when the operation time exceeds 10 seconds, it is considered that the amount of calculation exceeds the threshold. That is, when the amount of data is too large and the amount of calculation is too large, the multi-threaded architecture is enabled and additional threads are added for parallel calculation. At this time, the initialization and algorithm matching process of the path planning simulation system 12 are executed separately, which can reduce the operation time, accelerate the system operation, and thus improve the system's operation efficiency.
[0050] The initialization of the path planning simulation system 12 is to initialize some parameters of the demonstration so that the viewer can watch the demonstration process more clearly and directly. The initialization process specifically involves the initialization of the vehicle and its storage information, the customer point information and the corresponding distance matrix.
[0051] The data layer 3 of this embodiment preferably includes a system configuration database 31, an order information database 32 and an algorithm record database 33. The order information database 32 can be used to store data information records input by the operator, and the algorithm record database 33 can be used to store algorithm execution process records. Such a design can be used to check relevant data and records during review, thereby improving the stability and reliability of the system.
[0052] The vehicle routing problem (VRP-PD) is an extension of the VRP problem. It is an abstraction of the core problem of logistics distribution management - the problem of distribution vehicle scheduling. It can be described as: the vehicle starts from the station to serve the user, and returns to the station after completing the user's needs. It is stipulated that each user can only be served by one vehicle and only once. The optimization goal of the problem is to select an appropriate path so that the total path spent on completing all needs is minimized while satisfying the above constraints, thereby saving costs and increasing benefits.
[0053] The topological full sort traversal result can obtain the optimal solution, but it is very time-consuming. This embodiment proposes to optimize the allocation scheme under the constraint of topological relationship through a genetic algorithm based on topological relationship (GA-TR), using an approximate solution method suitable for the calculation of large-scale nodes, providing a high solution speed, which can greatly reduce the time complexity, so that the online algorithm can be better used in the actual scenario of logistics distribution. That is, this embodiment uses the genetic algorithm GA-TR based on topological relationship as the best algorithm for the VRP-pd problem. The GA-TR algorithm follows the traditional genetic algorithm process and produces offspring through selection (such as roulette), crossover (such as sequential crossover) and mutation. The algorithm is characterized by maintaining the constraints of topological relationships in each operation process.
[0054] Here, this embodiment only describes the parts that are different from the traditional genetic algorithm, such as Figure 2 As shown, the application process of the GA-TR algorithm of this embodiment mainly includes:
[0055] S01. Initialize the population: Randomly select path nodes with zero in-degree for sorting, maintaining the constraints of topological relationships.
[0056] The initial population mainly uses topological sorting to generate new individuals (single path solutions) to ensure the feasibility of the population (all path solutions). In specific applications, the path nodes with zero in-degree are randomly selected for sorting to ensure the diversity of the population. In the process of generating individuals, the requirements of the directed acyclic graph need to be met, and the starting and ending points of the chromosomes must be removed.
[0057] S02, crossover: Select two path solutions that conform to the topological sorting from the population for crossover.
[0058] Specifically, first select two paths parent1 and parent2 (these parents are called chromosomes in genetic algorithms) that meet the topological sorting from the population, and randomly intercept a part of the path gene (called gene in genetic algorithms) from parent1. Then traverse all nodes in parent2, and if there is the same part of the path in parent2, temporarily ignore it and skip the same part in parent2.
[0059] In addition, it is necessary to determine whether there is a pre-order (or post-order) node for the partial path gene. If there is a pre-order (or post-order) node, the pre-order (or post-order) node is added to the front end (or end) of the partial path gene. As for the remaining nodes, they are moved according to the sequential cross method.
[0060] S03, mutation: randomly select a path exchange node, determine the exchange range that meets the conditions, and exchange it with another path node randomly selected within the exchange range to generate a new path plan.
[0061] Due to the topological relationship, it is impossible to randomly exchange two nodes in the path. Therefore, first, randomly select an exchange node gene1. Secondly, it is necessary to determine the range of other exchange nodes. Starting from node gene1, find the first non-zero in-degree node and record its position in the path; then find the first non-zero out-degree node backward and record its position. In this way, the range of the mutation operation is determined, that is, from the first non-zero in-degree node to the first non-zero out-degree node. Finally, randomly select another node gene2 in the range and exchange gene1 and gene2. This mutation operation ensures that the new path satisfies the topological relationship.
[0062] S04. By repeating the above iterative process of crossover and mutation until the total distance of the new path plan is the shortest, the best plan can be output.
[0063] After a round of crossover and mutation is completed, by comparing the two topologically sorted path solutions used for crossover with the new path solution, if the total distance of the new path solution is not the shortest, continue to step S02, otherwise, output it as the best solution.
[0064] like Figure 3As shown, the demonstration interface of the simulation platform, namely the business application layer 1, is shown. The left area corresponds to the order entry system 11, presenting an interactive interface for inputting / selecting various parameters, and the right area corresponds to the path planning simulation system 12, presenting a demonstration interface for intuitive viewing. The operator selects a specific VRP problem according to the need, and the technical layer 2 matches the corresponding algorithm according to the VRP problem and selects to import the order instance. After starting the system, the algorithm obtains the optimal allocation strategy of the instance. In the demonstration interface on the right, the vehicle will complete its task, go from the departure area to the user's cargo storage location, load the cargo, and go to the corresponding order unloading area to unload the cargo. In this process, the total distance of the vehicle is continuously calculated for verification. For each vehicle, they need to complete their own tasks separately, starting from the storage location, going to the location of the corresponding order, and finally returning to the nearest warehouse address.
[0065] like Figure 4 As shown, the implementation method of the multi-vehicle task allocation and path optimization simulation platform of this embodiment mainly includes:
[0066] (1) Interact with the operator.
[0067] During the interaction, the operator inputs data information. The main steps of the interaction include: S10, select the VRP problem that needs to be visualized; S11, import customer point information (customer coordinates) into the system; S12, import vehicle warehouse information (vehicle and warehouse coordinates, quantity and load capacity, etc.) into the system; S13 import order demand information (goods distribution plan between users) into the system. When these data information are successfully imported, it can be determined that the interaction is completed (step S14), and the next step of algorithm calculation and demonstration can be carried out. Otherwise, the operator needs to re-enter the data information until the data information is successfully imported. From then on, the preprocessing step is completed.
[0068] (2) Select the best algorithm for the VRP problem from the system's built-in algorithms based on the data information input by the operator;
[0069] In order to speed up the operation of the simulation platform, in the process of selecting (matching) the best algorithm, this embodiment also determines in advance whether the process calculation amount for matching the best algorithm is huge (such as Figure 4In step S15), when the amount of calculation of the best algorithm matching process exceeds the threshold, the multi-threaded architecture is enabled (step S16), the best algorithm is obtained, and then the optimization strategy is calculated according to the best algorithm (step S17), that is, the optimal task allocation and path plan. For example, when the matching operation time exceeds 10 seconds, it is considered that the amount of calculation exceeds the threshold. That is, when the amount of data is too large and the amount of calculation is too large, the multi-threaded architecture loading algorithm is enabled, and additional threads are added for parallel calculation. At this time, the initialization and algorithm matching process of the path planning simulation system 12 are executed separately, which can reduce the operation time, speed up the system operation, and thus improve the system's operation efficiency.
[0070] The above are the steps to implement the VRP problem. In the specific simulation process, the simulation platform of this embodiment can also solve the problem of how to implement the extension of the VRP problem. The specific steps include: introducing the constraint parameter conditions of the extension problem based on the VRP problem in real time, and selecting the best algorithm for the extension problem adapted to it from the system built-in algorithm according to the data information input by the operator. For example, when the operator selects the VRP-pd problem, the GA-TR algorithm with the best performance will be automatically adapted.
[0071] (3) Demonstrate the logistics plan distribution process in real time according to the optimal algorithm (step S18).
[0072] After the optimization strategy is derived in the previous steps, the logistics tasks corresponding to the best algorithm can be assigned to the designated vehicles, and the following simulation demonstration can be carried out to show the physical picture to the viewers.
[0073] like Figure 5 As shown in the figure, when the simulation platform is performing a simulation demonstration, the simulation platform will first be initialized and wait for the matching algorithm to be executed. If the waiting time exceeds 10 seconds, the multi-threaded architecture will be enabled to expand the parallel computing of threads to increase the computing processing capability and save time. After the matching algorithm is executed, the best algorithm is transmitted to the vehicle, warehouse and customer information points. After the data of each information point is updated, the vehicle executes the task according to the best path planning plan. Every time it arrives at a customer point, the information data will be updated and it will be determined whether it is the last task. If it is, it will return to the nearest warehouse, otherwise it will continue to the next customer point.
[0074] The embodiments of the present invention aim at the VRP problem and its extension problems that need to be solved in the logistics environment, aim at providing a visual demonstration, and solve the shortcomings of the existing vehicle task allocation and path optimization application fields, such as the lack of pertinence, complex simulation process, poor scalability, and weak purpose. It mainly involves customer order processing, vehicle task allocation, and multi-vehicle path optimization. Customer order processing improves the efficiency of the logistics system by reasonably arranging the order of order completion and allocating vehicles; vehicle task allocation arranges the target and order of vehicle pickup by using a heuristic algorithm, and uses the total pickup and delivery distance as a measurement indicator. By optimizing the total delivery distance, the efficiency of the logistics system is improved; multi-vehicle path optimization presents the algorithm operation process and the logistics plan distribution process to the viewer in a visual way, so that the viewer can intuitively grasp the algorithm operation.
[0075] The advantages of the simulation platform and its implementation method of the embodiment of the present invention are mainly reflected in: first, the simulation platform designed for the logistics distribution system can display the operation process of the algorithm, and can provide a verification platform for various algorithms with the shortest distance as the optimization goal; second, combined with the visualization plug-in of the simulation system, a modular structure is adopted, which can provide algorithm verification for VRP and its extended problems, and has the characteristics of high flexibility and good scalability; third, the operator can modify the order quantity, the number of delivery vehicles and the load scale, etc., to realize simulation tests of VRP problems of different scales, so that the simulation platform has the characteristics of user customization. For example, the research problem can also be extended based on the existing one, such as adding the first-in-last-out constraint of the cargo task under VRP-pd, and studying new problems. It has certain flexibility and scalability and can quickly respond to sudden practical problems.
[0076] In summary, the multi-vehicle task allocation and path optimization simulation platform of the present invention has three layers: business application layer, technical layer, and data layer. After the operator selects a specific vehicle path problem according to the needs, the system matches the corresponding optimal algorithm according to the problem, and can perform a real-time demonstration of the vehicle's logistics plan distribution process according to the optimal algorithm, providing a visual verification platform for various algorithms with the shortest distance as the optimization goal. In addition, the VRP-pd problem is solved by using a genetic algorithm based on topological sorting, and an approximate solution method is used to provide a high solution speed, which can be applied online in actual scenarios. Moreover, when the calculation amount of the algorithm process is too large, the multi-threaded architecture can be enabled to accelerate the system operation.
[0077] The above is only a specific implementation of the present application. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principles of the present application. These improvements and modifications should also be regarded as the scope of protection of the present application.
Claims
1. A multi-vehicle task allocation and path optimization simulation platform, It is characterized in that include: Business application layer, including order entry system and route planning simulation system; The technical layer includes the algorithm matching module; The data layer includes a system configuration database, wherein the system configuration database stores a plurality of system built-in algorithms; The order entry system is used to interact with the operator, the algorithm matching module is used to select the best algorithm for the vehicle routing problem adapted to the operator from the system built-in algorithms according to the data information input by the operator, and the routing simulation system is used to demonstrate the logistics plan distribution process in real time according to the best algorithm; The order entry system interacts with the operator, including the selection of the vehicle routing problem to be visualized, and the pre-processing step of importing customer point information, order demand information and vehicle storage information into the system in real time; The technical layer further includes a research problem expansion module, which is used to introduce constraint parameter conditions of an extended problem based on the vehicle routing problem in real time, and the algorithm matching module is also used to select the best algorithm of the extended problem of the vehicle routing problem adapted thereto from the system built-in algorithms according to the data information input by the operator; Among them, the extended problem of the vehicle routing problem includes the loading and unloading vehicle routing problem, and the optimal algorithm of the vehicle routing problem is a genetic algorithm based on topological relations, which generates offspring through selection, crossover and mutation, and maintains the constraints of topological relations in each operation process; Wherein, the genetic algorithm based on topological relationship includes: Initialize the population: randomly select path nodes with zero in-degree and sort them, maintaining the constraints of topological relationships; Crossover: Select two path solutions that meet the topological sorting from the population for crossover; Mutation: Randomly select a path exchange node, determine the exchange range that meets the conditions, and exchange it with another path node randomly selected within the exchange range to generate a new path plan; Repeat the above iterative process until the total distance of the new path solution is the shortest.
2. The multi-vehicle task allocation and path optimization simulation platform according to claim 1, It is characterized in that The path planning simulation system is also used to display the data information input by the operator and the operation process of the algorithm in real time.
3. The multi-vehicle task allocation and path optimization simulation platform according to claim 1, It is characterized in that The technical layer also includes a data preprocessing module and a multi-threaded algorithm operation module. The data preprocessing module is used to determine whether the calculation amount of the matching best algorithm process is huge, and the multi-threaded algorithm operation module is used to enable the multi-threaded architecture when the calculation amount of the matching best algorithm process exceeds a threshold.
