A method and system for multi-AGV path planning and control in a production workshop

Abstract

This invention relates to a multi-AGV path planning and control method for production workshops, comprising: establishing a corresponding 0-1-2 grid map model according to the application scenario of the production workshop, wherein 0 represents passable roads, 1 represents AGV stopping points and goods storage and retrieval points, and 2 represents the space occupied by manufacturing equipment and tooling; performing path planning for multiple AGVs, including a bottom-level single AGV path planning algorithm and an upper-level path conflict retrieval and elimination algorithm, outputting a set of conflict-free paths for multiple AGVs; and controlling the working state of each AGV according to the set of conflict-free paths for multiple AGVs to complete the corresponding transportation tasks. Compared with the prior art, this invention improves the accuracy of map modeling, can more accurately describe the environmental information of the production workshop, and can quickly and accurately plan conflict-free paths for multiple AGVs, effectively improving the efficiency of goods transportation.

Description

Technical Field

[0001] This invention relates to the field of AGV path planning technology, and in particular to a multi-AGV path planning and control method and system for production workshops. Background Technology

[0002] With rising labor costs and technological advancements, traditional production workshops are transforming into unmanned and intelligent models. Intelligent logistics and distribution systems are a crucial component of intelligent unmanned workshops. These systems break down logistical barriers between warehouses and production lines, as well as various production processes, effectively improving logistics efficiency. Within these systems, AGVs (Automated Guided Vehicles) are the actual executors of material transportation, delivering materials to various stations promptly based on instructions from the intelligent logistics and distribution system.

[0003] In smart warehouses or production workshops, a single AGV is insufficient for efficient transportation; multiple AGVs typically operate simultaneously. To achieve intelligent material delivery, a workshop map model is needed to plan conflict-free paths for each AGV. Currently, different map modeling methods yield models with varying accuracy, which can impact the efficiency of subsequent path planning algorithms. For instance, grid map models are widely used in warehousing systems due to their ease of construction, representation, and unique location. However, in manufacturing systems, in addition to AGVs and the goods to be transported, there are obstacles such as machine tools and fixtures, making modeling more complex and requiring consideration of more factors. Furthermore, path planning is a low-cost and computationally intensive problem of finding a path sequence in the planning space based on prior map information and appropriate path planning algorithms. Essentially, it is about finding a conflict-free path from the initial point to the target point. In the case of a single AGV, path conflict is not a concern. In simple systems, when a small number of AGVs have conflicts, traffic rules can be used to eliminate them. However, in recent years, with the increase in the number of AGVs deployed in production workshops, the robot density has increased and the number of conflicts has increased. Traditional path conflict elimination methods have a large computational load and long solution time, which cannot meet the requirements of fast route planning in actual production and cannot ensure the accuracy of path planning. Summary of the Invention

[0004] The purpose of this invention is to overcome the shortcomings of the prior art and provide a multi-AGV path planning and control method and system for production workshops, which can quickly and accurately plan conflict-free paths for multiple AGVs.

[0005] The objective of this invention can be achieved through the following technical solution: a multi-AGV path planning and control method for a production workshop, comprising the following steps:

[0006] S1. Based on the application scenario of the production workshop, establish a corresponding 0-1-2 grid map model, where 0 represents passable roads, 1 represents AGV docking points and cargo storage and retrieval points, and 2 represents the space occupied by manufacturing equipment and tooling.

[0007] S2. Perform path planning for multiple AGVs, including the underlying single AGV path planning algorithm and the upper-level path conflict retrieval and elimination algorithm, and output the solution set of conflict-free paths for multiple AGVs.

[0008] S3. Based on the solution set of conflict-free paths for multiple AGVs, control the working status of each AGV accordingly to complete the corresponding transportation task.

[0009] Furthermore, step S2 specifically includes the following steps:

[0010] S21. Use the improved A* algorithm to plan the corresponding path for each AGV and obtain the initial path for multiple AGVs;

[0011] S22. For the initial path of multiple AGVs, first perform path conflict search. If a path conflict is found, proceed to step S23. Otherwise, output the initial path of multiple AGVs as a set of conflict-free paths.

[0012] S23. Based on the retrieved path conflicts, execute the path conflict elimination algorithm on the initial paths of multiple AGVs, and output the solution set of conflict-free paths for multiple AGVs.

[0013] Furthermore, in step S21, the improved A* algorithm specifically introduces a multi-minimum straight-line priority rule based on the A* algorithm. The evaluation function of the improved A* algorithm is f(P) = g(P) + h(P), where f(P) represents the estimated distance from the starting position S through the intermediate position P to the target position G, g(P) is the actual shortest path distance from the starting position S to the intermediate position P, and h(P) is the estimated path distance from the intermediate position P to the target position G. During the path planning process, when there are multiple equal minimum evaluation function values, the point that enables the AGV to travel in a straight line is selected as the path planning point.

[0014] Furthermore, the specific process of improving the A* algorithm to plan the corresponding path for a single AGV in step S21 is as follows:

[0015] S211, Input the workshop grid map GM, starting point S, and target point G;

[0016] S212. Initialize the OPEN table containing the starting position and the CLOSED table containing the empty set;

[0017] S213. Calculate f(S) = g(S) + h(S);

[0018] S214. Sort the OPEN table according to the f value, and select the point with the smallest f value as point P (if there are multiple equal minimum values, select the point that allows the AGV to travel in a straight line as point P). The first time, only the initial point S can be selected.

[0019] S215. Add the selected point P to the CLOSED table and check if node P is equal to the target point G. If it is, find the path; otherwise, execute S216 to S218.

[0020] S216, Search for the point P adjacent to P. adjacents If P adjacent If it is a conflict point or already in the CLOSE table, then do not process it, and continue searching for other neighboring points of P;

[0021] If P adjacent If it is not a conflict point and is not in the CLOSE table, then calculate f(P). adjacent );

[0022] S217, Check P adjacent Whether it is in the OPEN table, i.e., check P adjacent Has point P been searched? adjacent In the OPEN table, further checks are performed on f(P). adjacent Is ) less than f(P)? adjacent_open ), where P adjacent and P adjacent_open They represent the same point, f(P) adjacent ) represents the search range from the starting point to the ending point, passing through the intermediate point P. adjacent The estimated distance, f(P) adjacent_open ) is the P that was found in the previous search. adjacent From the starting point to the ending point and passing through the intermediate point P adjacent Estimated distance;

[0023] If f(P) adjacent )<f(P adjacent_open If P, then adjacent_open P adjacent replace;

[0024] If f(P) adjacent )≥f(P adjacent_open If P is maintained in the OPEN table, then P is maintained. adjacent_open ;

[0025] If P adjacent If not in the OPEN table, then set P. adjacent Insert into the OPEN table;

[0026] S218. Repeat steps S214 to S217 to obtain multiple path planning points in sequence. Connect the multiple path planning points to form the initial path of a single AGV.

[0027] Furthermore, the path conflict retrieval in step S22 includes edge conflict retrieval and vertex conflict retrieval. The vertex conflict retrieval detects whether two AGVs move to the same position within the same time step.

[0028] The edge conflict retrieval method detects whether two adjacent AGVs have swapped positions at the next time step, crossing the same edge in opposite directions.

[0029] Furthermore, the formula for determining vertex conflict retrieval is as follows:

[0030] a it (x p ,y p ) = a jt (x p ,y p )

[0031] Among them, a it (x p ,y p Let be the grid position p of the i-th AGV at time t, and let a be the position of the AGV at time t. jt (x p ,y p Let ) represent the grid position p of the j-th AGV at time t, when a it (x p ,y p ) = a jt (x p ,y p If the condition is met, then it is determined that there is a vertex conflict between the i-th AGV and the j-th AGV; otherwise, it is determined that there is no vertex conflict between the i-th AGV and the j-th AGV.

[0032] The formula for determining edge conflict retrieval is as follows:

[0033]

[0034] Among them, a it (x q ,y q Let ) represent the grid position q of the i-th AGV at time t, and a j(t+1) (x q ,y q Let ) represent the grid position q of the j-th AGV at time t+1, and a i(t+1) (x s ,y sLet be the grid position s of the i-th AGV at time t+1, and let a be the position of the AGV in grid s. jt (x s ,y s Let ) represent the grid position of the j-th AGV at time t;

[0035] when a it (x q ,y q ) = a j(t+1) (x q ,y q And a i(t+1) (x s ,y s ) = a jt (x s ,y s If all conditions are met, then it is determined that there is an edge conflict between the i-th AGV and the j-th AGV; otherwise, it is determined that there is no edge conflict between the i-th AGV and the j-th AGV.

[0036] Furthermore, step 23 specifically includes the following steps:

[0037] S231. Based on the path conflicts found, establish a conflict count set R.count, a conflict target set R.conf, and a conflict constraint set R.conset, where R... k .count represents AGV k The number of collisions with other AGVs;

[0038] R k .conf indicates the relationship with AGV k Conflicting sets of AGVs;

[0039] R k .conset represents AGV k The time and spatial location of conflicts with other AGVs.

[0040] S232. Based on the conflict quantity set R.count, conflict target set R.conf, and conflict constraint set R.conset, and combined with the Most Conflict First (MCF) rule, the Earliest Conflict First (ECF) rule, and the Single Search Conflict Point (SSCP) rule, eliminate the conflicts between AGVs. The Most Conflict First (MCF) rule means that when eliminating conflicts, the AGV with the most conflicts with other AGVs is selected first for conflict elimination.

[0041] The earliest conflict priority ECF rule adds the earliest conflict point C of the AGV that has eliminated the conflict to the constraint C.constraints for path replanning;

[0042] The Single Conflict Point (SSCP) rule is that if conflict point C was added to C.constraints during a previous AGV path replanning during a search process, and the earliest conflict point of a subsequent AGV is also conflict point C, the subsequent AGV will search for the subsequent conflict points of the earliest conflict point.

[0043] Furthermore, the specific process of step S232 is as follows:

[0044] Step 1: For the conflicting AGVs in the conflict count set R.count, sort them in descending order according to the number of conflicts between each AGV and other AGVs.

[0045] Step 2: Determine the number of AGVs to eliminate each time based on the number of conflicting AGVs (NCA);

[0046] Step 3: Select the AGV to be conflict-resolved and use the Most Conflicts First (MCF) rule, that is, when resolving conflicts, first select the AGV with the most conflicts with other AGVs for conflict resolution.

[0047] Step 4: Select the conflict point and use the earliest conflict first (ECF) rule to add the earliest conflict point C of the AGV that has eliminated the conflict to the constraint C.constraints.

[0048] Step 5: Considering constraints C.constraints, use the improved A* algorithm to replan a path for the conflict-resolved AGV;

[0049] Step 6: Repeat Steps 3-5 until the required number of conflict-resolved AGVs are reached;

[0050] Step 7: Calculate R.count, R.conf, and R.conset;

[0051] Step 8: Determine if the number of conflicts between AGVs is 0. If it is, the path conflict has been eliminated; otherwise, return to Step 1-Step 7.

[0052] A multi-AGV path planning and control system for a production workshop includes a grid map construction module and a path planning module, wherein the grid map construction module is used to build a grid map model of the production workshop;

[0053] The path planning module performs global path planning for multiple AGVs based on the grid map model of the production workshop, and plans conflict-free paths for all AGVs.

[0054] Compared with the prior art, the present invention has the following advantages:

[0055] I. This invention improves the accuracy of map modeling by establishing a 0-1-2 grid map model corresponding to the production workshop, thus more accurately describing the environmental information of the production workshop. For multiple AGVs, this invention designs a bottom-level single AGV path planning algorithm and an upper-level path conflict retrieval and elimination algorithm. Specifically, the bottom-level algorithm first plans an initial path for each AGV, and then the upper-level algorithm performs path conflict retrieval and elimination on the initial paths of multiple AGVs, thereby obtaining a conflict-free path solution set. This allows for fast and accurate planning of conflict-free paths for multiple AGVs, thereby improving transportation efficiency.

[0056] Second, in this invention, the underlying algorithm for multi-AGV path planning is an improved A* algorithm. That is, without considering other AGVs, the improved A* algorithm is used to plan the path of a single AGV by introducing multiple minimum value straight-line priority rules, which can better adapt to the needs of workshop path planning.

[0057] Third, in this invention, when performing path conflict retrieval and elimination, on the one hand, by retrieving vertex conflicts and edge conflicts, the conflicts existing in the initial paths of multiple AGVs can be fully and accurately retrieved. On the other hand, a conflict quantity set R.count, a conflict target set R.conf, and a conflict constraint set R.conset are introduced, and the rules of maximum conflict priority (MCF), earliest conflict priority (ECF), and single-search conflict point (SSCP) are set to eliminate the retrieved path conflicts. The proposed method has a small computational load, speeds up the solution time for conflict elimination, realizes rapid planning of multiple AGV paths, and improves the efficiency of material transportation in the production workshop. Attached Figure Description

[0058] Figure 1 This is a schematic diagram of the method flow of the present invention;

[0059] Figure 2 This is a schematic diagram illustrating the application process of the embodiment;

[0060] Figure 3 This is a schematic diagram of the 0-1-2 grid established in the embodiment;

[0061] Figure 4 This is a schematic diagram of vertex conflict in this invention;

[0062] Figure 5 This is a schematic diagram of edge conflict in the present invention. Detailed Implementation

[0063] The present invention will now be described in detail with reference to the accompanying drawings and specific embodiments.

[0064] Example

[0065] like Figure 1As shown, a multi-AGV path planning and control method for a production workshop includes the following steps:

[0066] S1. Based on the application scenarios in the production workshop, establish a corresponding 0-1-2 grid map model;

[0067] S2. Perform path planning for multiple AGVs, including the underlying single AGV path planning algorithm and the upper-level path conflict retrieval and elimination algorithm, and output the solution set of conflict-free paths for multiple AGVs.

[0068] S3. Based on the solution set of conflict-free paths for multiple AGVs, control the working status of each AGV accordingly to complete the corresponding transportation task.

[0069] This embodiment applies the above-described technical solution, such as Figure 2 As shown, the main contents include:

[0070] In step S1, establish as follows Figure 3 The 0-1-2 grid map shown represents a passable road, 1 represents AGV docking points and cargo storage / retrieval points, and 2 represents the space occupied by manufacturing equipment and tooling. For each time step, each AGV has four options: move to an adjacent location, wait in place, or turn left or right.

[0071] In step S2, when executing the underlying single AGV path planning algorithm, other AGVs are not considered. A multi-minimum straight-line priority rule is introduced, and an improved A* algorithm is used to plan the path of a single AGV. The evaluation function of the improved A* algorithm is f(P) = g(P) + h(P), where f(P) represents the estimated distance from the starting position S through the intermediate position P to the target position G, g(P) is the actual shortest path distance from the starting position S to the intermediate position P, and h(P) is the estimated path distance from the intermediate position P to the target position G. In this embodiment, the Manhattan distance is used, i.e., h(P) = |x P -x G |+|y P -y G In the path planning process for a single AGV, if multiple minimum evaluation function values ​​are equal, the point that allows the AGV to travel in a straight line is selected.

[0072] The specific process of the improved A* algorithm proposed in this scheme is shown below.

[0073] 2.1: Input the workshop grid map GM, starting point S, and target point G;

[0074] 2.2: Initialize the OPEN table containing the starting position and the CLOSED table containing the empty set;

[0075] 2.3: Calculate f(S) = g(S) + h(S);

[0076] 2.4: Sort the OPEN table according to the f value, and select the point with the smallest f value as point P (the initial point S can only be selected the first time);

[0077] 2.5: Add the selected point P to the CLOSED table and check if node P is equal to the target point G. If it is, find the path; otherwise, proceed to steps 2.6-2.8.

[0078] 2.6: Search for points P adjacent to P adjacents If P adjacent If it is a conflict point (considered during path conflict replanning) or already in the CLOSE table, it is not processed, and the algorithm continues to search for other neighboring points of P. If P adjacent If it is not a conflict point and is not in the CLOSE table, then calculate f(P). adjacent );

[0079] 2.7: Check P adjacent Whether it is in the OPEN table, i.e., check P adjacent Has point P been searched? adjacent In the OPEN table, further checks are performed on f(P). adjacent Is ) less than f(P)? adjacent_open ), where P adjacent and P adjacent_open They represent the same point, f(P) adjacent ) represents the search range from the starting point to the ending point, passing through the intermediate point P. adjacent The estimated distance, f(P) adjacent_open ) is the P that was found in the previous search. adjacent From the starting point to the ending point and passing through the intermediate point P adjacent Estimated distance;

[0080] If f(P) adjacent )<f(P adjacent_open If P, then adjacent_open P adjacent replace;

[0081] If f(P) adjacent )≥f(P adjacent_open If P is maintained in the OPEN table, then P is maintained. adjacent_open ;

[0082] If P adjacent If not in the OPEN table, then set P. adjacent Insert into the OPEN table;

[0083] 2.8: Perform steps 2.4-2.7.

[0084] In step S2, when executing the upper-level path conflict retrieval and elimination algorithm, path conflict detection is first performed, starting by determining vertex conflicts, such as... Figure 4 As shown, to determine whether two AGVs have moved to the same grid cell, formula a is used. it (x p ,y p ) = a jt (x p ,y p The equation is used to determine if there is a vertex conflict between the two AGVs.

[0085] Next, determine edge conflicts, such as Figure 5 As shown, to determine whether two adjacent AGVs have swapped positions and crossed the same edge in opposite directions at the next time step, the formula is used. If both equations are true, then an edge conflict is determined between the two AGVs.

[0086] Next, path conflict elimination is performed. Based on the conflicts found, a conflict count set R.count, a conflict target set R.conf, and a conflict constraint set R.conset are established, where R... k .count represents AGV k The number of collisions with other AGVs, R k .conf indicates the relationship with AGV k Conflicting AGV set, R k .conset represents AGV k The time and spatial location of conflicts with other AGVs.

[0087] Then perform the following steps, Steps 1-8, to eliminate the retrieved path conflicts:

[0088] Step 1: For the conflicting AGVs in the conflict count set R.count, sort them in descending order according to the number of conflicts between each AGV and other AGVs.

[0089] Step 2: Determine the number of AGVs to eliminate each time based on the number of conflicting AGVs (NCA).

[0090] Step 3: Select the AGV to be conflict-eliminating and adopt the Most Conflicts First (MCF) rule, that is, when eliminating conflicts, first select the AGV with the most conflicts with other AGVs for conflict elimination.

[0091] Step 4: Select the conflict point and use the earliest conflict priority ECF rule to add the earliest conflict point C of the AGV that has eliminated the conflict to the constraint C.constraints.

[0092] Step 5: Consider the constraints C.constraints and use the improved A* algorithm to replan a path for the conflict-free AGV.

[0093] Step 6: Execute Steps 3-5 until the required number of conflict-resolved AGVs are reached.

[0094] Step 7: Calculate R.count, R.conf, and R.conset.

[0095] Step 8: Determine if the number of conflicts between AGVs is 0. If it is, the algorithm is complete; otherwise, return to Step 1-Step 7.

[0096] In this embodiment, 50 AGVs in the manufacturing workshop simultaneously perform path planning. After conflict detection, the number of conflicting AGVs (NCA) is 32. This embodiment uses half the number of conflicting AGVs as the number of AGVs to eliminate in each conflict, i.e., NCA / 2. After sorting, the AGV with the most conflicts is [AGV Name]. 28 AGV 13 AGV 36 AGV 49 AGV 22 AGV7, ... Regarding AGVs 28 ,have:

[0097] A 28 .count=5

[0098] A 28 .conf = {AGV8, AGV 13 AGV 18 AGV 20 AGV 33}

[0099]

[0100] In A 28 In the .conset file, the first column represents the AGV number involved in the conflict, the second and third columns represent the location of the conflict, and the fourth column represents the time of the conflict. 28 The earliest conflict AGV was AGV 13 The conflict occurs at time 4. Therefore, position and time information (25, 34, 4) is added to the constraints, and the AGV is replanned. 28 The path. Then, replan the AGV. 13 The path. For AGV 13 ,have:

[0101] A 13 .count=4

[0102] A 13 .conf = {AGV4, AGV 28 AGV 36 AGV 45}

[0103]

[0104] AGV 28 During path replanning, the earliest conflict (25, 34, 4) was eliminated. According to the Single Search Conflict Point SSCP rule, if conflict point C was added to C.constraints during a previous AGV path replanning during a search, and the earliest conflict point for a subsequent AGV is also C, then the subsequent AGV will search for subsequent conflict points of the earliest conflict point. The earliest conflict after time 4 (25, 50, 21) was selected and added to C.constraints for the AGV. 13 Perform path replanning. Then, eliminate the selected conflicts one by one until there are no conflicts between the AGVs.

[0105] Therefore, it can be seen that the multi-AGV path planning method adopted in this solution for production workshops can efficiently and quickly plan conflict-free paths, improve logistics and transportation efficiency, and increase system response speed. It can provide a strong reference for application scenarios such as container terminal automation, drone traffic management, airport operation, and automated valet parking.

[0106] This embodiment also provides a multi-AGV path planning system for production workshops by applying the above-described method. This multi-AGV path planning system for production workshops can be used to implement the above-described multi-AGV path planning method for production workshops. Specifically, the multi-AGV path planning system for production workshops includes a grid map construction module and a path planning module.

[0107] The raster map building module is used to create a workshop raster map model;

[0108] The path planning module then performs global path planning for multiple AGVs based on the established workshop grid map model, planning a conflict-free path for all AGVs.

[0109] Furthermore, this embodiment provides a terminal device, which includes a processor and a memory. The memory stores a computer program, which includes program instructions. The processor executes the program instructions stored in the computer storage medium. The processor may be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. It is the computing and control core of the terminal, and is suitable for implementing one or more instructions. Specifically, it is suitable for loading and executing one or more instructions to achieve corresponding method flows or corresponding functions. The processor described in this embodiment can be used in the operation of a multi-AGV path planning method in a production workshop, including: S1, establishing a grid map model of the production workshop; S2, planning a conflict-free path for all AGVs.

[0110] This embodiment also provides a computer-readable storage medium (Memory), which is a memory device in a terminal device used to store programs and data. It is understood that the computer-readable storage medium here can include both the built-in storage medium in the terminal device and extended storage media supported by the terminal device. The computer-readable storage medium provides storage space that stores the terminal's operating system. Furthermore, this storage space also stores one or more instructions suitable for loading and execution by a processor; these instructions can be one or more computer programs (including program code). It should be noted that the computer-readable storage medium here can be high-speed RAM or non-volatile memory, such as at least one disk storage device.

[0111] One or more instructions stored in a computer-readable storage medium can be loaded and executed by the processor to implement the corresponding steps of the multi-AGV path planning method for production workshops in the above embodiments; one or more instructions in the computer-readable storage medium are loaded and executed by the processor to perform the following steps: S1, establish a grid map model; S2, plan a conflict-free path for all AGVs.

[0112] In summary, this technical solution first establishes a 0-1-2 grid map model, then performs path planning for AGVs. The underlying algorithm for path planning introduces a multi-minimum straight-line priority rule and uses an improved A* algorithm to plan the path of a single AGV. The upper-level algorithm searches for vertex and edge conflicts, establishing a conflict count set R.count, a conflict target set R.conf, and a conflict constraint set R.conset. It introduces the Most Conflict First (MCF) rule, the Earliest Conflict First (ECF) rule, and the Single Search Conflict Point (SSCP) rule to eliminate conflicts between AGVs. This improves the modeling accuracy of the production workshop and enables rapid and accurate planning of conflict-free paths for multiple AGVs, thereby improving the efficiency of goods transportation.

Claims

1. A multi-AGV path planning and control method for a production workshop, characterized in that, Includes the following steps: S1. Based on the application scenario of the production workshop, establish a corresponding 0-1-2 grid map model, where 0 represents passable roads, 1 represents AGV docking points and cargo storage and retrieval points, and 2 represents the space occupied by manufacturing equipment and tooling. S2. Perform path planning for multiple AGVs, including the underlying single AGV path planning algorithm and the upper-level path conflict retrieval and elimination algorithm, and output the solution set of conflict-free paths for multiple AGVs. S3. Based on the solution set of conflict-free paths for multiple AGVs, control the working status of each AGV accordingly to complete the corresponding transportation task. Step S2 specifically includes the following steps: S21. Use the improved A* algorithm to plan the corresponding path for each AGV and obtain the initial path for multiple AGVs; S22. For the initial path of multiple AGVs, first perform path conflict search. If a path conflict is found, proceed to step S23. Otherwise, output the initial path of multiple AGVs as a set of conflict-free paths. S23. Based on the retrieved path conflicts, execute the path conflict elimination algorithm on the initial paths of multiple AGVs, and output the solution set of conflict-free paths for multiple AGVs. Step S23 specifically includes the following steps: S231. Based on the path conflicts found, establish conflict count sets respectively. R.count Conflict target set R.conf and conflict constraint set R.conset ,in, R k .count for AGV k The number of collisions with other AGVs; R k .conf Indicates and AGV k Conflicting sets of AGVs; R k .conset express AGV k The time and spatial location of conflicts with other AGVs; S232, Based on the set of conflict counts R.count Conflict target set R.conf and conflict constraint set R.conset By combining the Most Conflicts First (MCF) rule, the Earliest Conflict First (ECF) rule, and the Single Search Conflict Point (SSCP) rule, conflicts between AGVs are eliminated.

2. The multi-AGV path planning and control method for a production workshop according to claim 1, characterized in that, The improved A* algorithm in step S21 specifically introduces a multi-minimum straight-line priority rule based on the A* algorithm. The evaluation function of the improved A* algorithm is: ,in, Indicates starting position S via the middle position P Reach the target location G The estimated distance, To find the starting position S To the middle position P The actual shortest path distance, From the middle position P To the target location G The estimated path distance is calculated. During the path planning process, if there are multiple cases where the minimum evaluation function values ​​are equal, the point that enables the AGV to travel in a straight line is selected as the path planning point.

3. The multi-AGV path planning and control method for a production workshop according to claim 2, characterized in that, The specific process of planning the corresponding path for a single AGV using the improved A* algorithm in step S21 is as follows: S211, Input workshop grid map GM ,starting point S Target point G ; S212, Initialize including the starting position OPEN Tables and empty sets CLOSED surface; S213, Calculation ; S214, according to f Value sorting OPEN table ,choose f The point with the smallest value is taken as the point. P The first time, only the initial point can be selected. S ; S215, Select the points P Put in CLOSED In the table, and check the nodes. P Is it equal to the target point? G If so, find the path; otherwise, execute S216~S218. S216, Search and P Adjacent points ,if Is it a point of conflict or already exists? CLOSE If the data in the table is not processed, the search continues. P Other adjacent points; if Not a point of conflict and not in CLOSE In the table, the calculation is performed. ; S217, Inspection Is it in OPEN The table shows the inspection Has the point been searched? exist OPEN The table then undergoes further examination. Is it less than ,in and They represent the same point. It is the search range from the starting point to the ending point, including intermediate points. The estimated distance, That was found in the previous search. From the starting point to the ending point and passing through the intermediate points. Estimated distance; like ,but quilt replace; like Then in OPEN Maintain in the table ; if Not here OPEN In the table, then Insert into OPEN In the table; S218. Repeat steps S214 to S217 to obtain multiple path planning points in sequence. Connect the multiple path planning points to form the initial path of a single AGV.

4. The multi-AGV path planning and control method for a production workshop according to claim 1, characterized in that, The path conflict retrieval in step S22 includes edge conflict retrieval and vertex conflict retrieval. The vertex conflict retrieval is to detect whether two AGVs move to the same position within the same time step. The edge conflict retrieval method detects whether two adjacent AGVs have swapped positions at the next time step, crossing the same edge in opposite directions.

5. A multi-AGV path planning and control method for a production workshop according to claim 4, characterized in that, The formula for determining vertex conflict retrieval is as follows: in, Let i be the grid cell where the i-th AGV is located at time t. p Location, Let j be the grid cell where the j-th AGV is located at time t. p Position, when If the condition is met, then it is determined that there is a vertex conflict between the i-th AGV and the j-th AGV; otherwise, it is determined that there is no vertex conflict between the i-th AGV and the j-th AGV. The formula for determining edge conflict retrieval is as follows: in, Let i be the grid cell where the i-th AGV is located at time t. q Location, The grid where the j-th AGV is located at time t+1 q Location, The grid where the i-th AGV is located at time t+1 s Location, Let j be the grid cell where the j-th AGV is located at time t. s Location; when and If both conditions are met, then it is determined that there is an edge conflict between the i-th AGV and the j-th AGV; otherwise, it is determined that there is no edge conflict between the i-th AGV and the j-th AGV.

6. The multi-AGV path planning and control method for a production workshop according to claim 1, characterized in that, The Most Conflict Priority (MCF) rule means that when eliminating conflicts, the AGV with the most conflicts with other AGVs is selected first for conflict elimination. The earliest conflict priority (ECF) rule prioritizes the earliest conflict point of the AGV that resolves the conflict. C Add constraints C.constraints Perform route replanning; The Single Search Conflict Point (SSCP) rule states that during a single search process, conflict points... C It was added during the previous AGV path replanning. C.constraints If the earliest point of conflict for the subsequent AGVs is also this point of conflict... C The AGV will then search for subsequent conflict points of the earliest conflict point.

7. A multi-AGV path planning and control method for a production workshop according to claim 6, characterized in that, The specific process of step S232 is as follows: Step 1: For the set of conflict numbers R.count The conflicting AGVs are sorted in descending order based on the number of conflicts between each AGV and other AGVs. Step 2: Determine the number of AGVs to eliminate each time based on the number of conflicting AGVs; Step 3: Select the AGV to be conflict-resolved and use the Most Conflicts First (MCF) rule, that is, when resolving conflicts, first select the AGV with the most conflicts with other AGVs for conflict resolution. Step 4: Select the conflict point and use the earliest conflict first (ECF) rule to assign the AGV that will resolve the conflict to the earliest conflict point. C Add constraints C.constraints ; Step 5: Consider Constraints C.constraints An improved A* algorithm is used to replan a path for the conflict-free AGV; Step 6: Repeat Steps 3-5 until the required number of conflict-resolved AGVs are reached; Step 7, Calculation R.count , R.conf , R.conset ; Step 8: Determine if the number of conflicts between AGVs is 0. If it is, the path conflict has been eliminated; otherwise, return to Step 1-Step 7.

8. A multi-AGV path planning control system applying the multi-AGV path planning control method for production workshops as described in claim 1, characterized in that, It includes a raster map construction module and a path planning module, wherein the raster map construction module is used to build a raster map model of the production workshop; The path planning module performs global path planning for multiple AGVs based on the grid map model of the production workshop, and plans conflict-free paths for all AGVs.

Images