AGV network path planning method and system of multi-agent path planning algorithm based on lazy constraint addition

Abstract

The invention provides an AGV network path planning method and system of a multi-agent path planning algorithm based on lazy constraint addition. The method comprises the steps of receiving input data, generating an initial path, extracting a non-deterministic node, optimizing a dynamic path, generating a Monte Carlo path, dynamically introducing an alternative path and outputting the path. Through optimization of an existing multi-agent path planning algorithm based on lazy constraint addition, the method not only can effectively plan a collision-free multi-AGV path in a complex environment, but also can dynamically adjust and optimize the path, and ensures that a nearly optimal solution is obtained in real-time operation. By dynamically introducing an alternative path, the flexibility and the response speed of path optimization are improved; a more uniformly distributed path is calculated in advance by using a space dispersion optimization technology, and conflicts and congestion among AGVs are reduced; in combination with a Monte Carlo configuration generation method, an optimal path is generated and screened for multiple times, and it is ensured that an optimal solution is found within limited time.

Description

Technical Field

[0001] The present application relates to the technical field of automatic guided vehicles, and in particular to an AGV network path planning method and system based on a multi-agent path planning algorithm with lazy constraint addition. Background Art

[0002] With the rapid development of Industry 4.0 and smart manufacturing, automated guided vehicles (AGVs) are increasingly used in modern warehousing, logistics, and manufacturing environments. AGVs move in complex industrial scenarios through preset paths or real-time planning to perform material transportation, loading and unloading, and scheduling tasks. In large warehouses or manufacturing plants, multiple AGVs often need to operate at the same time, and these AGVs must collaborate efficiently without collision in limited space and time. This poses a severe challenge to path planning algorithms, especially when dealing with multiple AGVs at the same time, where path conflicts, computational efficiency, and planning quality all need to be effectively resolved.

[0003] Traditional multi-agent path planning (MAPF) algorithms, such as A*, Dijkstra, and conflict-based search algorithms (CBS), perform well in small-scale scenarios. In large-scale, multi-agent scenarios, graph-based search algorithms such as A* and Dijkstra need to traverse all possible path combinations, resulting in an exponential increase in the amount of computation with the number of agents. This computational complexity often makes it difficult for these algorithms to obtain solutions within a reasonable time, and the computational performance is greatly reduced. At the same time, algorithms such as A* and Dijkstra perform well when dealing with a single agent, but when dealing with multiple agents, the planning complexity increases significantly due to the need to avoid collisions between different agents. These algorithms lack an effective conflict resolution mechanism, which leads to poor performance in dynamic and high-density environments; for CBS, conflicts become very common in large-scale scenarios, resulting in a large number of sub-problems after each split. These sub-problems will form a huge search tree. The rapid expansion of the search space causes the performance of conflict-based search algorithms (CBS) to significantly decrease when multiple conflicts need to be considered, and the efficiency of discrete conflict handling is even lower.

[0004] Therefore, there is an urgent need for an algorithm that can achieve efficient, collision-free, and near-optimal path planning in complex environments. Summary of the invention

[0005] The present application provides an AGV network path planning method and system based on a multi-agent path planning algorithm with lazy constraint addition, which can be used to solve the technical problem that the performance of traditional multi-agent path planning is significantly reduced when considering multiple conflicts.

[0006] The present application provides an AGV network path planning method based on a multi-agent path planning algorithm with lazy constraint addition, the method comprising the following steps:

[0007] Step 1, receiving input data: receiving the starting positions and target positions of multiple automated guided vehicles (AGVs), and converting them into a graphical model of a path planning problem, wherein the graphical model includes multiple vertices and edges;

[0008] Step 2, initial path generation: Based on the multi-agent path planning algorithm with lazy constraint addition, the initial solution is generated by calculating the shortest path of each AGV, where the cost of the path is based on the cumulative conversion cost from the starting position to the target position. The cumulative conversion cost is the sum of the number of moves of all agents when they have not reached the target state during the entire path planning; this measurement method is usually used to measure the quality of path planning. A smaller conversion cost means that the agent reaches the target in fewer steps or time, thereby achieving a better path planning solution.

[0009] Step 3, non-deterministic node extraction: In the process of path optimization, the non-deterministic node extraction method breaks the limitations of the traditional depth-first search mechanism, allowing the shortest path or fixed best node to be selected every time during the optimization phase. Instead, randomness is introduced to randomly select multiple feasible nodes for trial. This operation can prevent the search from falling into the local optimum, making the path selection more diversified, thereby reducing the probability of conflicts between different AGVs at the same node;

[0010] Step 4, dynamic path optimization: Dynamically optimize the initial path, use the spatially dispersed path optimization method, and pre-calculate some spatially dispersed paths to guide the path selection of each agent; in practical applications, these pre-calculated paths are distributed as evenly as possible in space and time, and the paths are dynamically adjusted to meet the goal of minimizing conflicts with other paths. The hyperparameter m is introduced to increase the search space and bypass congested areas. This method can effectively reduce conflicts between AGV paths and optimize path utilization by introducing non-shortest but more dispersed paths.

[0011] Step 5, Monte Carlo path generation: During the planning process, the Monte Carlo method is used to generate multiple candidate paths, and the optimal path is selected according to the cost function f(Q)=g(Q)+h(Q). The cost function is the sum of the cost g(Q) to the current point and the estimated cost h(Q) to the target. The cost function takes into account the length of the path, conflicts and time efficiency.

[0012] Step 6, dynamic introduction of alternative paths: In the path planning process, combined with the current solution, select some AGVs that may cause conflicts, such as intersections or narrow paths, and re-plan based on the solutions that have been found, or use local search or other path planning algorithms to re-plan, and find alternative paths that avoid conflicts, so as to further reduce path conflicts and optimize the movement efficiency of AGVs;

[0013] Step 7, path output: Output the final collision-free path set to the AGV control system to ensure that each AGV can drive to the target location according to the planned path.

[0014] Furthermore, the multi-agent path planning algorithm based on lazy constraint addition in the initial path generation in step 2 uses the algorithm characteristics at any time. When there are still unexplored nodes in the open stack or no interruption is encountered, the search and optimization will continue. In each search, the algorithm uses the heuristic function h to evaluate the distance between the current node and the target node, and compares it with the existing feasible solutions to find the potential optimal solution.

[0015] Furthermore, in step 2, the cumulative conversion cost formula is as follows:

[0016]

[0017] Where: Q t and Q t+1 are two consecutive configurations at time steps t and t+1, representing the set of positions of all agents;

[0018] cost e (Q t ,Q t+1 ): indicates that from configuration Q t To configure Q t+1 switching costs;

[0019] represents the number of agents that did not remain at the target position during the transition from configuration X to Y; that is, if agent i has not reached its target position g i , then it will contribute to the switching costs.

[0020] Furthermore, in step 3, non-deterministic node extraction is performed by randomly selecting a node in the search tree and selecting a node different from the top node of the current stack with a preset probability (e.g., 0.01), including two selection methods: a "restart" strategy of restarting the search from the starting node and a "random" strategy of randomly selecting a node in the stack. By introducing this method of non-deterministic node extraction, the search stagnation phenomenon that occurs during the path optimization process is avoided.

[0021] Furthermore, the dynamic path optimization in step 4 utilizes a spatially dispersed path optimization algorithm to pre-calculate several possible conflict-free shortest paths for each AGV and select a path set that minimizes total path conflicts to reduce path conflicts and congestion between AGVs.

[0022] Furthermore, step 5, Monte Carlo path generation includes:

[0023] Step 51, randomly generating a number of candidate paths multiple times;

[0024] Step 52, calculating the cost of each candidate path f(Q)=g(Q)+h(Q), where the cost is determined by factors such as path length, collision risk, and time overhead;

[0025] Step 53: Select the path with the minimum cost from the candidate paths as the final path.

[0026] Furthermore, step 6 is introduced, and the alternative path dynamics include:

[0027] Step 61, based on the current AGV path status, select AGVs that may cause conflicts, such as intersections or narrow paths, and generate local alternative paths;

[0028] Step 62, introducing an alternative path generated by a local optimization algorithm or other path planning algorithm;

[0029] Step 63, replace the existing path, escape from the local optimum, and re-explore a better path, thereby improving the quality of the solution.

[0030] The present application also provides an AGV network path planning system based on a multi-agent path planning algorithm with improved lazy constraint addition, the system is used to implement the method provided by the present application, and the system includes:

[0031] An input data receiving module is used to receive the starting positions and target positions of multiple automated guided vehicles (AGVs) and convert them into a graphic model of a path planning problem, where the graphic model includes multiple vertices and edges;

[0032] The initial path generation module is used for the multi-agent path planning algorithm based on lazy constraint addition. It generates an initial solution by calculating the shortest path of each AGV, where the cost of the path is based on the cumulative conversion cost from the starting position to the target position. The cumulative conversion cost is the sum of the number of moves of all agents when they have not reached the target state during the entire path planning; this measurement method is usually used to measure the quality of path planning. A smaller conversion cost means that the agent reaches the target with fewer steps or time, thereby achieving a better path planning solution.

[0033] The non-deterministic node extraction module is used to break the limitations of the traditional depth-first search mechanism through the non-deterministic node extraction method during the path optimization process. It allows the shortest path or the fixed best node to be selected every time during the optimization stage. Instead, it introduces randomness and randomly selects multiple feasible nodes for trial. This operation can avoid the search from falling into the local optimum, making the path selection more diversified, thereby reducing the probability of conflicts between different AGVs at the same node;

[0034] The dynamic path optimization module is used to dynamically optimize the initial path. It uses the spatially dispersed path optimization method to guide the path selection of each agent by pre-calculating some spatially dispersed paths. In practical applications, these pre-calculated paths are evenly distributed in space and time as much as possible, and the paths are dynamically adjusted to meet the goal of minimizing conflicts with other paths. The hyperparameter m is introduced to increase the search space and bypass congested areas. This method can effectively reduce conflicts between AGV paths and optimize path utilization by introducing non-shortest but more dispersed paths.

[0035] The Monte Carlo path generation module is used to generate multiple candidate paths using the Monte Carlo method during the planning process, and select the optimal path based on the cost function f(Q)=g(Q)+h(Q). The cost function is the sum of the cost g(Q) to the current point and the estimated cost h(Q) to the target. The cost function takes into account the length of the path, conflicts and time efficiency.

[0036] The alternative path dynamic introduction module is used to combine the current solution in the path planning process, select a part of the AGVs that may cause conflicts, such as intersections or narrow paths, and re-plan or use local search or other path planning algorithms to re-plan in the path planning process, so as to find alternative paths that avoid conflicts, further reduce path conflicts and optimize the movement efficiency of AGVs;

[0037] The path output module is used to output the final collision-free path set to the AGV control system to ensure that each AGV can drive to the target location according to the planned path.

[0038] Furthermore, the multi-agent path planning algorithm based on lazy constraint addition in the initial path generation module uses the algorithm characteristics at any time. When there are still unexplored nodes in the open stack or no interruption is encountered, the search and optimization will continue. In each search, the algorithm uses the heuristic function h to evaluate the distance between the current node and the target node, and compares it with the existing feasible solutions to find the potential optimal solution.

[0039] Furthermore, the cumulative conversion cost formula is as follows:

[0040]

[0041] Where: Q t and Q t+1 are two consecutive configurations at time steps t and t+1, representing the set of positions of all agents;

[0042] cost e (Q t ,Q t+1 ): indicates that from configuration Q t To configure Q t+1 switching costs;

[0043] represents the number of agents that did not remain at the target position during the transition from configuration X to Y; that is, if agent i has not reached its target position g i , then it will contribute to the switching costs.

[0044] The dynamic path optimization module is configured to optimize the paths of multiple AGVs through parallel computing, wherein the optimization process includes parallel computing of candidate paths, evaluating the cost of each path, and selecting a global optimal path.

[0045] The system can dynamically adjust the planned path according to changes in the actual environment, and use real-time generated alternative paths and optimization algorithms to ensure efficient collaboration of multiple AGVs and conflict-free paths.

[0046] This application aims to improve the path planning efficiency and optimization performance of AGV in complex dynamic environments, and solve the real-time, scalability and approximate optimality problems in AGV network path planning. Based on the research on multi-agent path planning (MAPF) problems, this invention proposes an improved multi-agent path planning algorithm based on lazy constraint addition (Lazy Constraint Addition for MAPF), and combines multiple optimization techniques to improve the initial solution quality and convergence speed of path planning, which is suitable for path planning of large-scale AGV clusters.

[0047] The multi-agent path planning algorithm based on lazy constraint addition is a search-based algorithm. Through the lazy constraint addition mechanism, after finding the initial solution, the quality of the solution is gradually improved, and finally converges to the optimal solution. The present invention has made many improvements on this basis, including: using non-deterministic search node extraction, effective use of pre-calculated decentralized paths, Monte Carlo successor generation methods, and dynamic introduction of alternative solutions. These improvements enable the improved multi-agent path planning algorithm based on lazy constraint addition to efficiently plan an approximate optimal path under large-scale, multi-constraint conditions. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] Figure 1It is a flow chart of the AGV path planning method of the present invention;

[0049] Figure 2 One of the schematic diagrams of the AGV path in the path planning process of the present invention;

[0050] Figure 3 This is the second schematic diagram of the AGV path in the path planning process of the present invention. DETAILED DESCRIPTION

[0051] In order to make the objectives, technical solutions and advantages of the present application clearer, the implementation methods of the present application will be further described in detail below with reference to the accompanying drawings.

[0052] In order to make the above-mentioned purposes, features and advantages of the present invention more obvious and easy to understand, the specific implementation methods of the present invention are described in detail below in conjunction with the drawings of the specification. Obviously, the described embodiments are part of the embodiments of the present invention, but not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary persons in the art without creative work should fall within the scope of protection of the present invention.

[0053] Example 1

[0054] In a multi-AGV task scenario, each AGV usually needs to perform multiple continuous transportation tasks in a complex environment. The present invention refines the task of each AGV by improving the multi-agent path planning algorithm based on lazy constraint addition, and decomposes the complex task into multiple subtasks based on the importance and time sequence of the task. Each subtask has a clear starting point and target point, indicating the process of the AGV moving from the starting point to the target point within a specific time period.

[0055] After the system receives the task information of multiple AGVs, it first uses the task decomposition module to split the overall task of each AGV into multiple subtasks. Each subtask corresponds to an independent path planning request to ensure the flexibility and accuracy of the task. The specific process is as follows:

[0056] S1. Initialization path planning: The system assigns an initial path to each AGV subtask, and uses pre-calculated dispersed paths to ensure that the initial path has good spatial distribution and reduce conflicts between AGVs.

[0057] S2. Initialize global constraints: including avoiding collisions between AGVs, maintaining a safe distance, and avoiding obstacles on the path.

[0058] S3. Set the key parameters of the improved multi-agent path planning algorithm based on lazy constraint addition, such as the adjustment coefficient of the heuristic function, the relaxation factor of the initial path, etc., to ensure that the algorithm can run efficiently in a complex environment.

[0059] By decomposing tasks and initializing path planning, the system provides a basic guarantee for subsequent dynamic adjustment and conflict resolution. This mechanism ensures that AGV can effectively complete its multiple transportation tasks and significantly reduce the probability of path conflicts.

[0060] Example 2

[0061] In practical applications, the working environment of AGV is often in a highly dynamic state. For example, a new transport task may be assigned while the AGV is executing an existing task. In addition, unexpected events on the path (such as the sudden appearance of obstacles or temporary failure of the AGV) may also affect the path planning and task execution of the AGV. Therefore, the present invention proposes a flexible path adjustment mechanism that can respond to these changes in real time and ensure the smooth completion of the AGV task. The specific implementation process is as follows:

[0062] S1. Real-time task redistribution: When a new task enters the system, the system will first evaluate the completion status of the existing task and the current position of the AGV. For those AGVs whose tasks are about to be completed, the system will directly assign them new tasks and quickly generate an optimized path from the end point of the task to the starting point of the new task based on the current position. For AGVs that have not yet completed their tasks, the system will determine the degree of conflict between the new task and the current path, and decide whether to adjust the existing path based on the conflict situation.

[0063] S2. Dynamic Path Response: This system has the function of dynamic local path adjustment. If a path conflict or environmental change (such as the appearance of a new obstacle) is detected during the execution of a task, the system will not replan the entire path, but will only adjust the affected local path. This real-time optimization of partial paths can improve the system response speed while reducing unnecessary computing overhead, allowing the AGV to quickly bypass obstacles or replan the path and continue to perform the current task.

[0064] S3. Task level and path priority adjustment: When the system detects certain high-priority tasks (such as urgent material delivery), it will adjust the priority of path planning according to the importance of the task. AGVs for important tasks will be given priority to shorter and less risky paths, while AGVs for other non-urgent tasks need to adjust their paths according to the new priority to avoid conflicts with high-priority AGVs. This task level priority mechanism ensures that high-priority tasks can be completed in a timely and efficient manner when multiple tasks are running in parallel.

[0065] This application aims to solve the problem of path conflicts when large-scale, multiple AGVs are running simultaneously, and significantly improve the efficiency of path planning and the quality of solutions. By optimizing the existing multi-agent path planning algorithm based on lazy constraint addition, this method can not only effectively plan collision-free multi-AGV paths in complex environments (such as large warehouses and manufacturing plants), but also dynamically adjust and optimize the paths to ensure that near-optimal solutions are achieved in real-time operation. The key technical improvements of the system include: improving the flexibility and response speed of path optimization by dynamically introducing alternative paths; using spatially dispersed optimization technology to pre-calculate more evenly distributed paths, reducing conflicts and congestion between AGVs; combining the Monte Carlo configuration generation method to generate and screen the best paths multiple times to ensure that the optimal solution is found within a limited time. This system can maintain efficient path planning and execution capabilities when dealing with up to hundreds of AGVs operating simultaneously, achieving real-time, collision-free, and near-optimal operation results.

[0066] The above-described embodiments of the present application do not constitute a limitation on the protection scope of the present application.

Claims

1. An AGV network path planning method based on a multi-agent path planning algorithm with lazy constraint addition, characterized in that: The method comprises the following steps: Step 1, receiving input data: receiving the starting positions and target positions of multiple automated guided vehicles (AGVs), and converting them into a graphical model of a path planning problem, wherein the graphical model includes multiple vertices and edges; Step 2, initial path generation: Based on the multi-agent path planning algorithm with lazy constraints, the initial solution is generated by calculating the shortest path for each AGV, where the cost of the path is based on the cumulative conversion cost from the starting position to the target position. The cumulative conversion cost is the sum of the number of moves of all agents that have not reached the target state during the entire path planning. Step 3, non-deterministic node extraction: In the optimization phase, instead of selecting the shortest path or a fixed optimal node every time, randomness is introduced to randomly select multiple feasible nodes for trial; Step 4, dynamic path optimization: Dynamically optimize the initial path, use the spatially dispersed path optimization method, and guide the path selection of each agent by pre-calculating some spatially dispersed paths; the pre-calculated paths are distributed as evenly as possible in space and time, and the paths are dynamically adjusted to meet the goal of minimizing conflicts with other paths. The hyperparameter m is introduced to increase the search space and bypass congested areas. Step 5, Monte Carlo path generation: Generate multiple candidate paths using the Monte Carlo method, and select the optimal path based on the cost function f(Q)=g(Q)+h(Q), where the cost function is the sum of the cost g(Q) to the current point and the estimated cost h(Q) to the target. Step 6, dynamic introduction of alternative paths: In the path planning process, combined with the current solution, select some AGVs that may cause conflicts, such as intersections or narrow paths, and re-plan them based on the solutions that have been found, or use local search or other path planning algorithms to re-plan, and find alternative paths that avoid conflicts; Step 7, path output: Output the final collision-free path set to the AGV control system to ensure that each AGV can drive to the target location according to the planned path.

2. The path planning method according to claim 1, characterized in that: The multi-agent path planning algorithm based on lazy constraint addition in the initial path generation step in step 2 uses the algorithm characteristics of any time. When there are still unexplored nodes in the open stack or no interruption is encountered, the search and optimization will continue. In each search, the algorithm uses the heuristic function h to evaluate the distance between the current node and the target node.

3. The path planning method according to claim 1, characterized in that: In step 2, the cumulative conversion cost formula is as follows: Where: Q t and Q t+1 are two consecutive configurations at time steps t and t+1, representing the set of positions of all agents; cost e (Q t ,Q t+1 ): indicates that from configuration Q t To configure Q t+1 switching costs; represents the number of agents that did not remain at the target position during the transition from configuration X to Y; that is, if agent i has not reached its target position g i , then it will contribute to the switching costs.

4. The path planning method according to claim 1, characterized in that: In step 3, non-deterministic node extraction is performed by randomly selecting a node in the search tree and selecting a node different from the top node of the current stack with a preset probability. There are two selection methods: a "restart" strategy that restarts the search from the starting node and a "random" strategy that randomly selects a node in the stack.

5. The path planning method according to claim 1, characterized in that: The dynamic path optimization in step 4 utilizes the spatially decentralized path optimization algorithm to pre-calculate several possible conflict-free shortest paths for each AGV and select the path set that minimizes the total path conflicts.

6. The path planning method according to claim 1, characterized in that: Step 5, Monte Carlo path generation includes: Step 51, randomly generating a number of candidate paths multiple times; Step 52, calculate the cost of each candidate path f(Q)=g(Q)+h(Q); Step 53: Select the path with the minimum cost from the candidate paths as the final path.

7. The path planning method according to claim 1, characterized in that: Introducing step 6, the alternative path dynamics include: Step 61, based on the current AGV path status, select AGVs that may cause conflicts, such as intersections or narrow paths, and generate local alternative paths; Step 62, introducing an alternative path generated by a local optimization algorithm or other path planning algorithm; Step 63, replace the existing path, escape from the local optimum, and re-explore a better path.

8. An AGV network path planning system based on a multi-agent algorithm with lazy constraint addition, the system being used to implement the method of any one of claims 1 to 7, characterized in that: The system comprises: An input data receiving module is used to receive the starting positions and target positions of multiple automated guided vehicles (AGVs) and convert them into a graphic model of a path planning problem, where the graphic model includes multiple vertices and edges; The initial path generation module is used for the multi-agent path planning algorithm based on lazy constraint addition. It generates the initial solution by calculating the shortest path of each AGV. The cost of the path is based on the cumulative conversion cost from the starting position to the target position. The cumulative conversion cost is the sum of the number of moves of all agents that have not reached the target state during the entire path planning. The non-deterministic node extraction module is used to introduce randomness and randomly select multiple feasible nodes for trial instead of selecting the shortest path or the fixed best node every time during the optimization phase; The dynamic path optimization module is used to dynamically optimize the initial path. It uses the spatially dispersed path optimization method to guide the path selection of each agent by pre-calculating some spatially dispersed paths. The pre-calculated paths are distributed as evenly as possible in space and time. The paths are dynamically adjusted to minimize conflicts with other paths. The hyperparameter m is introduced to increase the search space and bypass congested areas. A Monte Carlo path generation module is used to generate multiple candidate paths using the Monte Carlo method and select the optimal path according to the cost function f(Q)=g(Q)+h(Q), where the cost function is the sum of the cost g(Q) to the current point and the estimated cost h(Q) to the target; The alternative path dynamic introduction module is used to combine the current solution in the path planning process, select a part of the AGVs that may cause conflicts, such as intersections or narrow paths, and re-plan or use local search or other path planning algorithms to re-plan in the path planning process, and find alternative paths to avoid conflicts; The path output module is used to output the final collision-free path set to the AGV control system to ensure that each AGV can drive to the target location according to the planned path.

9. The system according to claim 8, characterized in that The multi-agent path planning algorithm based on lazy constraint addition in the initial path generation module uses the algorithm characteristics of any time. When there are still unexplored nodes in the open stack or no interruption is encountered, it continues to perform search and optimization. In each search, the algorithm uses the heuristic function h to evaluate the distance between the current node and the target node.

10. The system according to claim 8, characterized in that The cumulative conversion cost formula is as follows: Where: Q t and Q t+1 are two consecutive configurations at time steps t and t+1, representing the set of positions of all agents; cost e (Q t ,Q t+1 ): indicates that from configuration Q t To configure Q t+1 switching costs; represents the number of agents that did not remain at the target position during the transition from configuration X to Y; that is, if agent i has not reached its target position g i , then it will contribute to the switching costs.

Images