Method and device for planning a joint agricultural machine operation path

Abstract

The application discloses a joint agricultural machine operation path planning method and device, comprising: dividing a target farmland into multiple operation rows based on agricultural machine parameters and farmland parameters; assigning the multiple operation rows to multiple agricultural machines to be operated for joint operation in the target farmland; and determining an optimal joint agricultural machine operation path based on transfer costs of the agricultural machines to be operated under preset constraint conditions, wherein the transfer costs include an entering cost of each agricultural machine to be operated from a corresponding starting point of a warehouse to enter a corresponding assigned operation row, a row changing cost of each agricultural machine to be operated when operating through all corresponding assigned operation rows, and a leaving cost of each agricultural machine to be operated when returning to the corresponding starting point after completing operation, and the optimal joint agricultural machine operation path corresponds to a path of the multiple agricultural machines to be operated from the corresponding starting point of the warehouse, through all corresponding assigned operation rows, and returning to the corresponding starting point.

Description

Technical Field

[0001] This application relates to the field of path planning technology, and in particular to a method and apparatus for combined agricultural machinery operation path planning. Background Technology

[0002] In modern agricultural production, path planning is a key technology for modern intelligent agricultural machinery to achieve autonomous navigation and operation. It determines the total workload of agricultural machinery and affects the actual operational efficiency and quality. As farmland operations gradually become large-scale standardized farmland, researching multi-machine joint operations and scientifically allocating and scheduling resources can improve production efficiency.

[0003] Existing methods for planning the operation paths of combined agricultural machinery mostly only study the operation paths of the machinery in the field and the transfer paths between the operation rows, ignoring the specific transfer paths of the machinery entering or leaving the field. This leads to insufficient accuracy of the planned paths and reduces the environmental adaptability and safety of the machinery's autonomous operation. Summary of the Invention

[0004] The purpose of this application is to provide a method and apparatus for planning the operation path of combined agricultural machinery, so as to solve the problem of insufficient accuracy in the existing combined agricultural machinery operation path planning.

[0005] To solve the above-mentioned technical problems, this specification is implemented as follows:

[0006] Firstly, a method for planning the path of combined agricultural machinery operations is provided, including:

[0007] Based on agricultural machinery parameters and farmland parameters, the target farmland is divided into multiple work rows;

[0008] Multiple work rows are assigned to multiple agricultural machines waiting to be operated in conjunction with the target farmland;

[0009] Under preset constraints, the optimal combined agricultural machinery operation path is determined based on the transfer cost of each agricultural machine to be operated. The transfer cost includes the entry cost of each agricultural machine to be operated starting from the starting point corresponding to its warehouse and entering the corresponding assigned work row, the line-changing cost when performing operations through all the corresponding assigned work rows, and the departure cost when returning to the corresponding starting point after completing the operation. The optimal combined agricultural machinery operation path corresponds to the path of the multiple agricultural machines to be operated starting from the starting point corresponding to their warehouse, passing through all the corresponding assigned work rows, and returning to the corresponding starting point.

[0010] Optionally, dividing the target farmland into multiple work rows based on agricultural machinery parameters and farmland parameters includes:

[0011] Based on the farmland parameters of the target farmland, the longest boundary among the boundaries of the target farmland is selected as the baseline;

[0012] Parallel to the baseline, the machine is moved at equal intervals according to the machine operation width in the machine parameters to generate multiple operation rows until the entire target farmland is covered.

[0013] Optionally, determining the optimal combined agricultural machinery operation path based on the transfer cost of each agricultural machine to be operated under preset constraints includes:

[0014] The starting point of each agricultural machine waiting to be operated and the multiple operation rows assigned to each agricultural machine are taken as nodes;

[0015] The transition costs between nodes are stored in the adjacency matrix. The transition costs include the entry cost and exit cost between the starting node and the corresponding working row node, as well as the line break cost between the working row nodes.

[0016] The node traversal order that minimizes the transfer cost is determined by a preset global search algorithm;

[0017] Based on the node traversal order that minimizes the transfer cost, the optimal operation path for a single agricultural machine to be operated is determined.

[0018] The optimal combined agricultural machinery operation path is determined based on the optimal operation path of each agricultural machine to be operated.

[0019] Optionally, the step of assigning multiple work rows to multiple agricultural machines waiting to work in the joint operation on the target farmland includes: assigning multiple work rows to multiple agricultural machines waiting to work in the joint operation on the target farmland using the outer simulated annealing algorithm of the two-layer simulated annealing algorithm;

[0020] The step of determining the node traversal order with the lowest transfer cost for each agricultural machine under the assigned work row by using a preset global search algorithm includes: determining the node traversal order with the lowest transfer cost by using the inner simulated annealing algorithm of the two-layer simulated annealing algorithm.

[0021] Optionally, the outer simulated annealing algorithm, which uses a two-layer simulated annealing algorithm, assigns multiple job rows to multiple agricultural machines awaiting operation in the target farmland, including:

[0022] The number and location of the multiple work rows in the target farmland, as well as the number of multiple agricultural machines to be operated, are taken as input and provided to the outer simulated annealing algorithm to obtain the outer solution output by the outer simulated annealing algorithm. The outer solution includes the number and location of the work rows assigned to each agricultural machine to be operated.

[0023] Optionally, determining the node traversal order that minimizes the transfer cost for each agricultural machine under its assigned work row using the inner simulated annealing algorithm of the two-layer simulated annealing algorithm includes:

[0024] The outer layer solution is passed to the inner layer simulated annealing algorithm, which is used to calculate the inner layer solution and the corresponding transfer cost at different inner layer temperatures based on the outer layer solution.

[0025] The current inner layer temperature is lowered to the target inner layer temperature, and the traversal order of the work rows assigned to each agricultural machine to be operated in the inner layer solution is adjusted to obtain an updated inner layer solution and the corresponding transition cost. The inner layer solution includes the traversal order of the work rows assigned to each agricultural machine to be operated corresponding to the outer layer solution. The target inner layer temperature includes one of the inner layer temperatures that gradually decrease from the inner layer initial temperature preset by the inner layer simulated annealing algorithm and the inner layer termination temperature preset by the inner layer simulated annealing algorithm.

[0026] Based on the difference in transition costs before and after the inner solution is updated, the historically optimal inner solution is updated, and the optimal inner solution corresponds to the minimum transition cost.

[0027] If the target inner layer temperature reaches the inner layer termination temperature, the historical best inner layer solution is output; otherwise, the step of reducing the current inner layer temperature to the target inner layer temperature is returned until the inner layer termination temperature is reached.

[0028] The optimal combined agricultural machinery operation path is determined based on the traversal order of the nodes corresponding to the work rows of the historical best inner layer solution and the number and position of the work rows assigned to the outer layer solution.

[0029] Optionally, before determining the optimal combined agricultural machinery operation path based on the traversal order of the nodes corresponding to the historical best inner-layer solution and the number and position of the assigned work rows corresponding to the outer-layer solution, the method further includes:

[0030] Based on the transition cost corresponding to the historical best inner layer solution and the penalty term corresponding to the outer layer solution, the transition cost corresponding to the outer layer solution of the outer simulated annealing algorithm is determined.

[0031] The current outer layer temperature is lowered to the target outer layer temperature, and the number and position of the work rows assigned to each agricultural machine to be operated in the outer layer solution are adjusted to obtain an updated outer layer solution. The target outer layer temperature includes one of the outer layer temperatures that are gradually reduced from the outer layer initial temperature preset by the outer layer simulated annealing algorithm and the outer layer termination temperature preset by the outer layer simulated annealing algorithm.

[0032] Based on the updated outer solution, return to the step of passing the outer solution to the inner simulated annealing algorithm, until the inner simulated annealing algorithm outputs an updated historical best inner solution;

[0033] Based on the transition cost corresponding to the updated historical best inner layer solution and the penalty term corresponding to the outer layer solution, the transition cost corresponding to the updated outer layer solution is determined;

[0034] Based on the difference in transition costs before and after the outer solution is updated, the historically optimal outer solution is updated, and the optimal outer solution corresponds to the minimum transition cost.

[0035] If the target outer layer temperature reaches the outer layer termination temperature, the historical best outer layer solution is output; otherwise, the step of reducing the current outer layer temperature to the target outer layer temperature is returned until the outer layer termination temperature is reached.

[0036] The determination of the optimal combined agricultural machinery operation path based on the traversal order of the nodes corresponding to the historical best inner layer solution and the number and position of the assigned work rows corresponding to the outer layer solution includes:

[0037] The optimal combined agricultural machinery operation path is determined based on the traversal order of the nodes corresponding to the output historical best inner layer solution and the number and position of the assigned operation rows corresponding to the output historical best outer layer solution.

[0038] Optionally, it also includes:

[0039] In the optimal combined agricultural machinery operation path, if there is a conflict in the switching paths of two agricultural machines corresponding to the working rows, then an obstacle circle with the width of the agricultural machine as the diameter is set between the two starting points of the two agricultural machines in their respective switching paths.

[0040] Determine the target path points of the two agricultural machines on the obstacle circle, with the maximum perpendicular distance between the two target path points and the line connecting the two starting points;

[0041] Based on the starting point and path points of the two agricultural machines, the detour paths of the two agricultural machines on the switching path are determined by the Dubins path algorithm.

[0042] Optionally, the target agricultural machine to be operated is assigned to multiple work rows, and the row switching cost corresponding to the target agricultural machine to be operated is related to different turning methods. The row switching cost includes one or more of the following: distance cost, time cost, and reserved turning distance generated by the corresponding turning method.

[0043] In a second aspect, a combined agricultural machinery operation path planning device is provided, including a processor and a memory, wherein the memory stores a program or instructions that can run on the processor, and the program or instructions, when executed by the processor, implement the steps of the method described in the first aspect.

[0044] In this embodiment, the target farmland is divided into multiple work rows based on agricultural machinery parameters and farmland parameters. These work rows are then assigned to multiple agricultural machines scheduled to work together on the target farmland. Under preset constraints, the optimal combined agricultural machinery operation path is determined based on the transfer costs of each machine. These transfer costs include the entry cost for each machine starting from its designated warehouse and entering its assigned work row, the row-changing cost when working through all assigned work rows, and the departure cost when returning to its designated starting point after completing its work. The optimal combined agricultural machinery operation path corresponds to the path taken by the multiple machines starting from their designated warehouse, passing through all assigned work rows, and returning to their designated starting point. This allows for the comprehensive planning of the optimal combined agricultural machinery operation path by integrating the overall costs generated by the combined operation paths of the machines in the farmland, the transfer paths between work rows, and the specific transfer paths of the machines entering or leaving the farmland. This improves the accuracy of the planned path and enhances the environmental adaptability and safety of the autonomous operation of the agricultural machinery. Attached Figure Description

[0045] The accompanying drawings, which are included to provide a further understanding of this application and form part of this application, illustrate exemplary embodiments and are used to explain this application, but do not constitute an undue limitation of this application. In the drawings:

[0046] Figure 1 This is a flowchart illustrating the combined agricultural machinery operation path planning method according to an embodiment of this application.

[0047] Figure 2 This is a schematic diagram of the turning method for line breaks in the work row according to an embodiment of this application.

[0048] Figure 3 This is a schematic diagram of the combined agricultural machinery operation path planning in an embodiment of this application.

[0049] Figure 4 This is a schematic diagram of multiple agricultural machinery path conflicts according to an embodiment of this application.

[0050] Figure 5 This is a schematic diagram of multiple agricultural machinery path detours according to an embodiment of this application.

[0051] Figure 6 This is a schematic diagram of the overall process of the combined agricultural machinery operation path planning method according to an embodiment of this application.

[0052] Figure 7This is a structural block diagram of the combined agricultural machinery operation path planning device according to an embodiment of this application. Detailed Implementation

[0053] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application. The drawing numbers in this application are only used to distinguish the various steps in the solution and are not used to limit the execution order of the various steps. The specific execution order is subject to the description in the specification.

[0054] To address the problems existing in the prior art, embodiments of this application provide a method for planning the combined agricultural machinery operation path, such as... Figure 1 As shown, the process includes steps 102 to 106.

[0055] Step 102: Based on the agricultural machinery parameters and farmland parameters, the target farmland is divided into multiple work rows.

[0056] Agricultural machinery needs to cover the entire farmland area when operating, and it performs parallel operations by repeatedly traversing the farmland in rows. Key parameters for agricultural machinery include the machine's working width, while farmland parameters include the farmland's boundaries and length.

[0057] Based on the solution provided in the above embodiments, optionally, in step 102 above, dividing the target farmland into multiple work rows based on agricultural machinery parameters and farmland parameters includes: selecting the longest boundary of the target farmland as a baseline based on the farmland parameters of the target farmland; and translating the farmland parallel to the baseline at equal intervals according to the agricultural machinery operation width in the agricultural machinery parameters to generate multiple work rows until the entire target farmland is covered.

[0058] For farmland with regular boundary areas, a baseline translation method can be used. Using a baseline parallel to the longer boundary of the farmland, multiple straight working paths can be generated at equal intervals according to the working width of the agricultural machinery to be working in the farmland until the entire farmland area is covered, resulting in multiple working rows corresponding to the working width of the agricultural machinery.

[0059] Step 104: Assign multiple work rows to multiple agricultural machines waiting to work in the target farmland.

[0060] In this embodiment of the application, the scenario is applied to the joint operation of multiple agricultural machines. Therefore, it is necessary to assign the divided multiple work rows to the multiple agricultural machines to be operated in the corresponding joint operation.

[0061] Based on the arrangement and combination of different agricultural machines and different work rows, multiple different allocation schemes can be obtained. For example, if there are two agricultural machines waiting to be used in the target farmland, machine 1 and machine 2; and there are 6 work rows, which are numbered sequentially as row 1, row 2, ... row 6, then allocation scheme 1 can be obtained as follows: machine 1 is allocated to rows 1-3, and machine 2 is allocated to rows 4-6; allocation scheme 2 can be as follows: machine 1 is allocated to rows 1 and 3-4, and machine 2 is allocated to rows 2 and 5-6, and so on.

[0062] Step 106: Under preset constraints, determine the optimal combined agricultural machinery operation path based on the transfer cost of each agricultural machine to be operated. The transfer cost includes the entry cost of each agricultural machine to be operated starting from the starting point corresponding to its warehouse and entering the corresponding assigned work row, the line-changing cost when performing operations through all the corresponding assigned work rows, and the departure cost when returning to the corresponding starting point after completing the operation. The optimal combined agricultural machinery operation path corresponds to the path of the multiple agricultural machines to be operated starting from the starting point corresponding to their warehouse, passing through all the corresponding assigned work rows, and returning to the corresponding starting point.

[0063] After the allocation of work rows to multiple agricultural machines in step 104 is determined, the joint operation path corresponding to each allocation scheme can be determined. For example, the operation path of agricultural machine No. 1 in the joint operation path corresponding to allocation scheme 1 is as follows: Agricultural machine No. 1 starts from the starting point a corresponding to its warehouse, enters the first allocated work row, moves to the second row after completing the work in the first row, moves to the third row after completing the work in the second row, and returns to the corresponding starting point a after completing the work in the third row; or, agricultural machine No. 1 starts from the starting point a corresponding to its warehouse, enters the third allocated work row, moves to the second row after completing the work in the third row, moves to the first row after completing the work in the second row, and returns to the corresponding starting point a after completing the work in the first row. Correspondingly, the operation path of agricultural machine No. 2 in the joint operation path corresponding to allocation scheme 1 can be determined.

[0064] If the target agricultural machine is assigned an odd number of work rows, it needs to move to a designated path (the machine path is a road where the machine does not perform any work) when returning to the warehouse after completing its work. If the machine path is not used, and the machine returns to the warehouse via another machine's work row, it will cause conflicts with the work paths of other machines. Therefore, in this embodiment, the machine path can be used to return to the warehouse, thus solving the problem of an odd number of work rows being assigned by increasing the number of machine paths.

[0065] For example, the machine bay lane can be configured to be assigned to row 0 for the corresponding agricultural machine. Taking the above-mentioned agricultural machine No. 1 as an example, the normally assigned work rows are rows 1 to 3, so a machine bay lane needs to be added to assign it to rows 0 to 3. Correspondingly, the work rows assigned to agricultural machine No. 2 are row 0 and rows 4 to 6.

[0066] By analogy, the combined agricultural machinery operation path corresponding to each allocation scheme can be determined. Under preset constraints, the transfer costs generated by the combined agricultural machinery operation paths corresponding to different allocation schemes are different. The transfer costs include the entry cost for each agricultural machine to enter the corresponding allocated operation line from its starting point corresponding to its warehouse, the line-changing cost when performing operations through all corresponding allocated operation lines, and the departure cost when returning to the corresponding starting point after completing the operation.

[0067] The closer the determined combined agricultural machinery operation path of the allocation scheme is to meeting the preset constraints, the lower the corresponding transfer cost. Preset constraints include, for example, minimizing the distance of the combined agricultural machinery operation path, minimizing the combined agricultural machinery operation time, and minimizing the turning distance reserved when the combined agricultural machinery changes between different work rows. The combined agricultural machinery operation path with the minimum transfer cost is the optimal combined agricultural machinery operation path for each agricultural machinery to be operated. Multiple agricultural machinery to be operated start from the starting point corresponding to their respective warehouse, reach their assigned work row according to the optimal combined agricultural machinery operation path, perform their work, and return to their corresponding starting point after completing their assigned work row, thus minimizing the transfer cost.

[0068] In one embodiment, determining the optimal combined agricultural machinery operation path based on the transfer cost of each agricultural machine to be operated under preset constraints includes: using the starting point of each agricultural machine to be operated and the multiple work rows allocated to each agricultural machine as nodes; storing the transfer costs between nodes in an adjacency matrix, wherein the transfer costs correspond to the entry cost, exit cost, and row-changing cost between nodes in the work row and the node at the starting point; determining the node traversal order with the minimum transfer cost using a preset global search algorithm; determining the optimal operation path of each agricultural machine to be operated based on the node traversal order with the minimum transfer cost; and determining the optimal combined agricultural machinery operation path based on the optimal operation paths of each agricultural machine to be operated.

[0069] To meet the need for multiple agricultural machines to automatically navigate and traverse parallel work rows, this application embodiment considers the cost of a single agricultural machine entering and leaving the farmland. The work rows of the entire farmland are divided and assigned to each agricultural machine. The problem of optimizing the joint agricultural machine operation path is redescribed as dividing the joint agricultural machine operation path into multiple sub-operation paths and assigning them to individual agricultural machines.

[0070] For a single agricultural machine, the goal is to find the optimal path for the machine to start from its origin, traverse all assigned work rows, and return to its origin under various constraints. Since the paths within each work row are fixed, the work path optimization problem is essentially about minimizing the cost of switching between different work rows during the work process. By treating the origin and each work row as nodes and storing the transition costs between nodes in an adjacency matrix, a search algorithm can be used to find the work row traversal order with the minimum transition costs.

[0071] Adjacency matrix For a size The square formation, This represents the number of work rows assigned to a single target agricultural machine. If the traversal order of the work rows is... Then the transfer cost and ,in These are the unique job row node numbers. This is the starting node.

[0072] The feasible solution for the traversal order is the full permutation of the N work row nodes. By traversing each feasible solution, the global optimal solution is found. Based on the node traversal order corresponding to the global optimal solution, the optimal operation path of a single agricultural machine is determined, and then the optimal joint operation path of multiple agricultural machines is obtained.

[0073] As the number of nodes in the job row increases, the time cost of traversing all feasible solutions becomes significant. In one embodiment, this application employs a heuristic algorithm to search for the globally optimal solution.

[0074] Specifically, the step of assigning multiple work rows to multiple agricultural machines waiting to work in the target farmland includes: assigning multiple work rows to multiple agricultural machines waiting to work in the target farmland using the outer simulated annealing algorithm of the two-layer simulated annealing algorithm; the step of determining the node traversal order with the minimum transfer cost using a preset global search algorithm includes: determining the node traversal order with the minimum transfer cost for each agricultural machine waiting to work under the assigned work row using the inner simulated annealing algorithm of the two-layer simulated annealing algorithm.

[0075] In the above embodiments, the inner layer of the two-layer simulated annealing algorithm is used to search for the optimal row traversal order for a single agricultural machine, and then the outer layer of the two-layer simulated annealing algorithm is used to optimize the overall operation plan for multiple agricultural machines. Simulated annealing (SA) is derived from the principle of solid-state annealing and was first proposed by N. Metropolis et al. The simulated annealing algorithm accepts inferior solutions with a certain probability, enabling it to escape local optima and obtain the global optimum.

[0076] Specifically, the outer simulated annealing algorithm, which uses a two-layer simulated annealing algorithm, assigns multiple work rows to multiple agricultural machines waiting to work in the target farmland. This includes: taking the number and position of the multiple work rows divided in the target farmland, and the number of the multiple agricultural machines waiting to work, as inputs to the outer simulated annealing algorithm to obtain the outer solution output by the outer simulated annealing algorithm. The outer solution includes the number and position of the work rows assigned to each agricultural machine waiting to work.

[0077] In this embodiment, the outer simulated annealing algorithm is first initiated. Based on the number and location of the work rows in the farmland and the number of agricultural machines involved in the joint operation, the outer simulated annealing algorithm randomly generates an outer solution, global_solution1. This outer solution includes the allocation scheme for each agricultural machine to be operated, namely, the work row location and the number of crop rows assigned to each machine. For example, the randomly generated allocation scheme is: machine 1 is assigned to rows 0-3, and machine 2 is assigned to rows 0 and 4-6.

[0078] Specifically, the step of determining the node traversal order where the transfer cost is minimized for each agricultural machine under its assigned work row using the inner-layer simulated annealing algorithm of the two-layer simulated annealing algorithm includes: passing the outer-layer solution to the inner-layer simulated annealing algorithm, which calculates the inner-layer solution and corresponding transfer cost at different inner-layer temperatures based on the outer-layer solution; lowering the current inner-layer temperature to the target inner-layer temperature; and adjusting the traversal order of the work rows assigned to each agricultural machine in the inner-layer solution to obtain an updated inner-layer solution and corresponding transfer cost. The inner-layer solution includes the traversal order of the work rows assigned to each agricultural machine corresponding to the outer-layer solution, and the target inner-layer temperature... The temperature includes one of the following: the inner layer temperature gradually decreasing from the preset initial temperature of the inner layer simulated annealing algorithm and the preset termination temperature of the inner layer simulated annealing algorithm; based on the difference in the transition cost corresponding to the inner layer solution before and after the update, the historically optimal inner layer solution is updated, and the optimal inner layer solution corresponds to the minimum transition cost; if the target inner layer temperature reaches the inner layer termination temperature, the historically optimal inner layer solution is output; otherwise, the step of reducing the current inner layer temperature to the target inner layer temperature is returned; based on the traversal order of the nodes of the work row corresponding to the output historically optimal inner layer solution and the number and position of the work rows allocated to the outer layer solution, the optimal combined agricultural machinery operation path is determined.

[0079] In this embodiment, the outer simulated annealing algorithm passes the outer solution global_solution1 to the inner simulated annealing algorithm for optimization. The initial temperature of the inner simulated annealing algorithm is set. Cooling rate and stopping temperature Based on the job row positions and numbers allocated in the outer solution global_solution1, the inner solution at the corresponding inner temperature is calculated using the inner simulated annealing algorithm. This refers to the traversal order of the work rows assigned to each agricultural machine, and the transition cost corresponding to the assigned traversal order of the work rows. .

[0080] Inner layer simulated annealing algorithm can start from the initial temperature The temperature is gradually decreased, for example, at preset temperature intervals, with the target inner layer temperature being the temperature after each decrease. After each temperature decrease, an adjustment factor is calculated based on the current inner layer temperature T2. Solve from the inner layer Random selection The elements are randomly swapped, that is, the inner solution is randomly adjusted. The traversal order of the corresponding work rows assigned to each agricultural machine to be worked, for example, the inner solution The corresponding traversal order of the work rows is as follows: Farm machine 1 traverses the work rows in the order of row 1, row 2, row 3, row 0; Farm machine 2 traverses the work rows in the order of row 4, row 0, row 5, row 6. The traversal order of the work rows after random adjustment is as follows: Farm machine 1 traverses the work rows in the order of row 2, row 0, row 3, row 1; Farm machine 2 traverses the work rows in the order of row 5, row 6, row 4, row 0.

[0081] Adjust inner layer solution of After obtaining each element, a new inner-layer solution is obtained using the inner-layer simulated annealing algorithm. and its costs Among them, the adjustment factor , This is the floor factor. and These represent the maximum and minimum number of job rows that can be adjusted, respectively.

[0082] Calculate the transition cost before and after updating the inner solution. The difference between them, if Superior ,Right now , indicating inner solution If the corresponding transfer cost is smaller, then the inner solution is accepted. And update the historical best inner layer solution. Otherwise, by probability Accept new inner solution Simultaneously with the historical best inner solution In comparison, if The corresponding transfer cost is less than The corresponding transfer cost is then updated. for The goal is to minimize the transfer cost of maintaining the historically optimal inner-layer solution.

[0083] Determine if the current inner layer temperature T2 has dropped to the set stop temperature. If so, output the historical best inner solution. This yields the optimal traversal order of the work rows for each agricultural machine assigned to the corresponding outer layer solution. Therefore, based on the traversal order of the nodes corresponding to the work rows of the historical optimal inner layer solution and the number and position of the work rows assigned to the outer layer solution, the optimal combined agricultural machine operation path for each agricultural machine to be operated is determined.

[0084] If the current inner layer temperature T2 does not decrease to the set stop temperature Then return to the steps of gradually reducing the inner layer temperature, and repeat the above steps in a loop until the current inner layer temperature drops to the set stopping temperature. .

[0085] In one embodiment, the historical best inner layer solution output by the inner layer simulated annealing algorithm at different temperatures can be further analyzed. The corresponding transition cost is passed to the outer simulated annealing algorithm for optimization.

[0086] Specifically, before determining the optimal combined agricultural machinery operation path based on the traversal order of the nodes corresponding to the work rows of the output historical best inner layer solution and the number and position of the work rows allocated to the outer layer solution, the method further includes: determining the transfer cost corresponding to the outer layer solution of the outer layer simulated annealing algorithm based on the transfer cost corresponding to the output historical best inner layer solution and the penalty term corresponding to the outer layer solution; reducing the current outer layer temperature to the target outer layer temperature and adjusting the number and position of the work rows allocated to each agricultural machinery to be operated in the outer layer solution to obtain an updated outer layer solution, wherein the target outer layer temperature includes one of the outer layer temperatures that gradually decrease from the preset initial temperature of the outer layer simulated annealing algorithm and the preset termination temperature of the outer layer simulated annealing algorithm; and returning to the step of passing the outer layer solution to the inner layer simulated annealing algorithm based on the updated outer layer solution until the inner layer simulated annealing algorithm outputs an updated historical best inner layer solution. The optimal inner layer solution is determined; based on the transition cost corresponding to the updated historical best inner layer solution and the penalty term corresponding to the outer layer solution, the transition cost corresponding to the updated outer layer solution is determined; based on the difference in the transition costs corresponding to the outer layer solution before and after the update, the historical best outer layer solution is updated, and the optimal outer layer solution has the minimum transition cost; if the target outer layer temperature reaches the outer layer termination temperature, the historical best outer layer solution is output; otherwise, the step of reducing the current outer layer temperature to the target outer layer temperature is returned until the outer layer termination temperature is reached; wherein, the determination of the optimal combined agricultural machinery operation path based on the traversal order of the nodes of the operation row corresponding to the output historical best inner layer solution and the number and position of the operation rows allocated to the outer layer solution includes: determining the optimal combined agricultural machinery operation path based on the traversal order of the nodes of the operation row corresponding to the output historical best inner layer solution and the number and position of the operation rows allocated to the output historical best outer layer solution.

[0087] In the above embodiments, when the current inner layer temperature T2 is determined to have decreased to the set temperature, a stop temperature is reached. Output the historical best inner layer solution Then, the inner simulated annealing algorithm will use the historical best inner solution. The transition cost and its corresponding transfer cost are passed to the outer simulated annealing algorithm. The outer simulated annealing algorithm determines the transition cost corresponding to the current outer solution based on the transition cost and the penalty term corresponding to the current outer solution's allocation scheme (e.g., a certain penalty term is generated if the job rows in the allocation are unbalanced). And will be based on the historical best inner layer solution As the historical best outer layer solution of the outer layer simulated annealing algorithm at the current temperature .

[0088] In this embodiment, the outer simulated annealing algorithm passes the outer solution global_solution1 to the inner simulated annealing algorithm for optimization. The initial temperature of the outer simulated annealing algorithm is set. and stopping temperature .

[0089] The outer simulated annealing algorithm can be derived from the initial temperature. The temperature is gradually decreased, for example, at preset temperature intervals, with the target outer layer temperature being the temperature after each decrease. After each temperature decrease, an adjustment factor is calculated based on the current outer layer temperature T1. Decipher from the outer layer Random selection The elements are randomly swapped, that is, the outer solution is randomly adjusted. The corresponding allocation scheme for each agricultural machine to be operated, for example, the outer layer... The corresponding allocation scheme is as follows: Farm machine No. 1 is allocated to rows 0-3, and farm machine No. 2 is allocated to rows 0 and 4-6. The scheme is then randomly adjusted to: Farm machine No. 1 is allocated to rows 0, 2, 4, and 6, and farm machine No. 2 is allocated to rows 0, 1, 3, and 5.

[0090] Adjust outer layer solution After obtaining the first element, a new outer solution is obtained using the outer simulated annealing algorithm. and its costs Among them, the adjustment factor , This is the floor factor. and These represent the maximum and minimum number of job rows that can be adjusted, respectively.

[0091] Calculate the transition cost before and after updating the outer solution. The difference between them, if Superior ,Right now , indicating the outer solution If the corresponding transfer cost is smaller, then the outer solution is accepted. And update the historical best inner layer solution. Otherwise, by probability Accept new outer solution Simultaneously with the historical best outer layer solution In comparison, if The corresponding transfer cost is less than The corresponding transfer cost is then updated. for The goal is to minimize the transfer cost of maintaining the historically optimal outer layer solution.

[0092] Determine if the current outer layer temperature T1 has dropped to the set temperature stop temperature. If so, output the historical best outer solution. This yields the allocation scheme corresponding to the outer layer solution, and the optimal traversal order of the work rows allocated to each agricultural machine corresponding to the inner layer solution under the current outer layer temperature. Therefore, based on the traversal order of the nodes corresponding to the work rows of the historical best inner layer solution and the number and position of the work rows allocated to the outer layer solution, the optimal combined agricultural machine operation path for each agricultural machine to be operated is determined.

[0093] If the current outer layer temperature T1 does not decrease to the set stopping temperature Then return to the above steps of gradually decreasing the temperature, and repeat the above steps in a loop until the current outer layer temperature drops to the set stopping temperature. .

[0094] By using a two-layer simulated annealing search algorithm to search for the work row allocation scheme and traversal order of each agricultural machine, the search process is simplified and the planning efficiency is improved.

[0095] In the above embodiments, the transition cost calculated by the inner simulated annealing algorithm is determined based on the entry cost, line break cost, and exit cost of the path corresponding to the traversal order of the assigned work rows for each agricultural machine. The entry cost for each agricultural machine to enter its assigned work row from its starting point corresponding to its warehouse and the exit cost to return to its corresponding starting point after completing its work can be determined based on the path length determined by the Dubins path algorithm. When the path length determined by the Dubins path algorithm is the shortest, the corresponding entry cost or exit cost is minimized.

[0096] In two-dimensional Euclidean space, for two given points, their velocity vectors, and the minimum radius of curvature... The reasonable shortest path has the following possible cases: ① radius is An arc, followed by a line segment, and then a circle with a radius of... The arc is represented by CLC (circle-line-circle); ② The radius is The sequence of three circular arcs is denoted by CCC (circle-circle-circle); ③ is a sub-path of a CLC or CCC type path. This is the famous Dubins path problem.

[0097] For radius An arc can rotate clockwise or counterclockwise, and its rotation direction can be represented by the motion primitives right turn (R) and left turn (L). For a line segment, its direction is represented by the motion primitive straight (S). Therefore, a CCC-type path can be specifically represented by two combinations of motion primitives: LRL and RLR; a CLC-type path can be specifically represented by four combinations of motion primitives: LSL, RSR, RSL, and LSR. Based on the Dubins path algorithm, a local path planning method can be implemented. This method can plan paths from a given starting position. Departure speed Finish line Final speed and minimum radius of curvature Under certain conditions, the paths corresponding to the six optimal path combinations are planned, and the optimal path that satisfies the constraints is selected, which is the shortest path.

[0098] Given the known starting positions of each agricultural machine to be operated. The starting speed of each agricultural machine waiting to be used is the corresponding warehouse starting point. The agricultural machinery's travel speed and destination position. The initial entry position of the assigned work line, and the final speed. For agricultural machinery travel speed and minimum radius of curvature Given the turning radius of agricultural machinery, the Dubins path algorithm can be used to calculate the entry cost for each agricultural machine to enter the corresponding assigned work row from the starting point corresponding to its warehouse.

[0099] Similarly, the Dubins path algorithm can be used to calculate the departure cost for each agricultural machine to return to its corresponding starting point after completing its task.

[0100] The cost of changing lanes when agricultural machinery is operating in its assigned work row can be determined based on different turning patterns. Common turning patterns include the following four: bow shape, pear shape, fishtail shape, and hook shape. Different turning patterns have different path requirements.

[0101] Optionally, the target agricultural machine to be operated is assigned to multiple work rows, and the row switching cost corresponding to the target agricultural machine to be operated is related to different turning methods. The row switching cost includes one or more of the following: distance cost, time cost, and reserved turning distance generated by the corresponding turning method.

[0102] exist Figure 2 In the middle, use This indicates the minimum turning radius of the agricultural machinery. This indicates the number of work rows for the interval when turning. For example, if the target agricultural machine is assigned to row 1 and row 3, then the work row for the interval when turning from row 1 to row 3 is 1 row. Indicates the working width of agricultural machinery. This indicates the angle between the direction of travel of the agricultural machinery when it turns to leave or enter the current work row and the boundary of the farmland it is facing. Indicates the linear operating speed of agricultural machinery. θ represents the turning speed of the agricultural machinery, and θ represents the angle between the direction of travel of the agricultural machinery when turning away from or into the current working row and the direction of the current working row. A, B, C, and D represent the points passed in sequence during the turn.

[0103] Bow-shaped turning method Figure 2 As shown in (a), its distance cost (the length of the path required to turn) is The time cost (the time required to turn) is The reserved turning distance (the maximum length occupied by the agricultural machinery when turning) is .

[0104] Pear-shaped turning method Figure 2 As shown in (b) above, its distance cost is The time cost is The reserved turning distance is ,in .

[0105] Fishtail-shaped turning method Figure 2 As shown in (c), its distance cost is The time cost is The reserved turning distance is , This indicates the reversing speed. Since the fishtail turn involves reversing, its cost is generally lower than the pear-shaped turn, however, it places higher demands on the performance of the agricultural machinery.

[0106] Hook-type turning method such as Figure 2 As shown in (d) in the figure, its distance cost is ,in The time cost is The reserved turning distance is .

[0107] To facilitate searching for the optimal row traversal order, a row-swapping cost matrix (adjacency matrix) is constructed based on the desired turning pattern, representing the row-swapping costs between different rows. For a single agricultural machine, if the total number of assigned rows is N, then the corresponding adjacency matrix is ​​an N*N matrix denoted as H, where the elements... Indicates from the work line To the work line The cost of line breaks. Its turning radius. Interval with work lines The relationship is There is a line break when there is one line break. When established, one can choose between pear-shaped, fishtail-shaped, and bow-shaped turning patterns; when there is a gap of 2 rows or more, there is... Established, choose the bow-shaped turning method.

[0108] Based on different turning methods, the minimum turning cost of each agricultural machine to be operated can be calculated when changing between different operating rows, and the turning method corresponding to the minimum turning cost is determined as the optimal turning path for each agricultural machine to be operated.

[0109] Therefore, based on the entry cost of each agricultural machine to be operated from the starting point corresponding to its warehouse to enter the corresponding assigned work row, the line-changing cost when operating through all the corresponding assigned work rows, and the departure cost when returning to the corresponding starting point after completing the operation, the optimal combined agricultural machine operation path for each agricultural machine to be operated can be determined.

[0110] In this application example, by introducing the adjacency matrix of the agricultural machinery's line-changing cost and the Dubins algorithm path, the performance indicators corresponding to the various turning methods of the agricultural machinery and the paths for entering and leaving the farmland can be quantified, which can improve the calculation accuracy of path planning.

[0111] For example Figure 3 This is a schematic diagram of the combined agricultural machinery operation path planning according to an embodiment of this application. In this example, there are 4 agricultural machines 100 involved in the combined operation. Each agricultural machine 100 starts from the starting point corresponding to its warehouse, passes through all the corresponding assigned work rows, and returns to the corresponding starting point. The path is as follows: Figure 3 Different colors are used to indicate different tasks. Triangles represent entering a task row, and pentagrams represent leaving a task row.

[0112] After planning the optimal combined operation path for each agricultural machine to be operated, it is necessary to consider the conflicts that may occur when the agricultural machines change rows during the operation. Therefore, it is necessary to set up mutual obstacle avoidance for the agricultural machines to ensure the safety of the operation.

[0113] In one embodiment, the method further includes: if there is a conflict in the switching paths of two agricultural machines corresponding to the working rows in the optimal combined agricultural machinery operation path, then setting an obstacle circle with the width of the agricultural machine as the diameter between the two starting points in the respective switching paths of the two agricultural machines; determining the target path points of the two agricultural machines on the obstacle circle respectively, with the vertical distance between the two target path points and the line connecting the two starting points being the largest; and determining the detour path corresponding to the two agricultural machines on the switching path using the Durbins path algorithm based on the starting points and path points corresponding to the two agricultural machines.

[0114] like Figure 4As shown, when two agricultural machines encounter a path conflict while changing lanes in different work rows, for example, when the planned paths 10 and 20 have overlapping forward paths P1, if the forward path P1 is long enough, an obstacle avoidance path needs to be designed so that the two agricultural machines can alternately bypass the obstacle. If the forward path P1 is insufficient, due to the limited width of the farmland, the two agricultural machines cannot simultaneously bypass the obstacle. In this case, the agricultural machine with lower priority may need to pause its movement before leaving its current work row, while the agricultural machine with higher priority continues to move along its original planned path until the agricultural machine with higher priority bypasses the obstacle, at which point the agricultural machine with lower priority can continue its movement.

[0115] If the starting points of the corresponding work rows for both agricultural machines are known and the conflict path length is sufficient, such as Figure 5 As shown, when a farm machine on path 10 leaves its corresponding work row 12 and begins to go straight, the starting point of path P2 is a1. When another farm machine on path 20 leaves its corresponding work row 22 and begins to go straight, the starting point of path P2 is a2.

[0116] To achieve a detour, an obstacle circle Q with the width of the farm machinery as its diameter needs to be set between the two starting points a1 and a2. The two points on obstacle circle Q with the largest vertical distance L2 from the line L1 connecting the two starting points a1 and a2 are designated as waypoints b1 and b2. The line connecting starting points a1 and a2 should pass through the center of obstacle circle Q. Thus, the two points on obstacle circle Q furthest from the line L1 connecting starting points a1 and a2 are the top and bottom points on obstacle circle Q. After determining the starting points a1 and a2 and the waypoints b1 and b2 for each of the two farm machines, the detour path for the two farm machines on the switching path P1 is determined using the Dubins path algorithm. For one farm machine, starting point a1 corresponds to the starting position in the Dubins path algorithm, and waypoint b1 corresponds to the ending position. For the other farm machine, starting point a2 corresponds to the starting position in the Dubins path algorithm, and waypoint b2 corresponds to the ending position. Therefore, the optimal detour route that satisfies the constraints can be determined by using the Dubins path.

[0117] like Figure 5 As shown, when conflict paths occur, two agricultural machines start from different work rows and enter other work rows. When more agricultural machines work together, the lengths of the forward paths P1 of different pairs of conflicting agricultural machines may be different. Therefore, it is necessary to adjust the position of the obstacle circle Q and the starting point of the detour accordingly to determine whether mutual detours can be achieved and to find the optimal detour plan.

[0118] Once the length of the conflicting forward path P1 is obtained, it is only necessary to determine the starting points a1 and a2 of the two vehicles' respective detours. Then, the obstacle circle Q can be placed at the midpoint between the starting points a1 and a2, thereby obtaining the Dubins detour paths of the two agricultural machines and the paths they take to enter the new work row.

[0119] The determination of the starting points a1 and a2 can be based on a genetic algorithm search scheme. For ease of subsequent writing, the process of determining the detour path based on the starting points a1 and a2 will be defined as follows: , where x1 and x2 represent two starting points a1 and a2, respectively.

[0120] Individuals are represented as vectors ,in ( (where P1 is the length of the forward path). The population size is N1, the number of generations is G, and the mutation rate is μ. The steps of the genetic algorithm are as follows.

[0121] (a) Initialize the population:

[0122]

[0123] in and yes Random numbers on the array.

[0124] (b) For each generation Perform the following steps:

[0125] b1. Calculate fitness: for each individual Calculate its fitness:

[0126]

[0127] b2. Use roulette wheel selection method. Individual The probability of being selected is:

[0128]

[0129] b3. Crossover operation: Cross over the two selected parent individuals and Generate offspring , ,in yes Random numbers on the array.

[0130] b4. Mutation operation: For each gene With probability Mutation:

[0131]

[0132] in It is a small variance value.

[0133] if ,but ;if ,but .

[0134] b5. Elite Preservation: Retaining the most fit individuals from the previous generation.

[0135]

[0136] b6. Formation of a new generation of population:

[0137]

[0138] Termination condition: when The algorithm terminates when the time is right.

[0139] (d) Output result: Optimal solution .

[0140] Among them, superscript Indicates the first Subscript Indicates the first Individual, This represents the elite individuals of the g-th generation. and Let represent the independent variables that maximize and minimize the objective function, respectively. Based on the optimal output solution, the corresponding... This determines the starting points a1 and a2 for each of the two agricultural machines. Based on this, and considering the length of the conflicting forward path P1, the obstacle circle Q can be placed at the midpoint between starting points a1 and a2. This allows us to obtain the optimal Dubins bypass path for both machines and the path they take to enter the new working row.

[0141] If no suitable detour route can be found, the above priority strategy should be adopted to make the two agricultural machines move alternately to avoid path conflicts and improve the safety of agricultural machinery operations.

[0142] Figure 6 This is a schematic diagram of the overall process of the combined agricultural machinery operation path planning method according to an embodiment of this application. In this example, the two-layer simulated annealing search algorithm and the Dubins path algorithm are combined to take into account the agricultural machinery allocation scheme, the traversal order of the operation row and the obstacle avoidance of agricultural machinery conflict paths, so as to achieve the optimal combined agricultural machinery operation path.

[0143] like Figure 6 As shown, it includes the following steps:

[0144] Step 202: Based on the agricultural machinery parameters, farm (i.e., farmland) parameters, and the turning method of agricultural machinery row switching, determine the turning cost matrix, work row parameters, and number of agricultural machinery;

[0145] Step 204: Based on the parameters determined in step 202, and combined with a specific search method, such as the outer simulated annealing search algorithm of the two-layer simulated annealing search algorithm, search for the allocation scheme of the sub-operation rows corresponding to each agricultural machine in the combined agricultural machinery.

[0146] Step 206: Based on the allocation scheme determined in step 204 and the inner simulated annealing search algorithm of the two-layer simulated annealing search algorithm, search for the allocation operation row traversal order with the minimum turning cost of a single agricultural machine, wherein the optimal results searched by the outer simulated annealing search algorithm and the outer simulated annealing search algorithm are mutually optimized.

[0147] Step 208: Combine agricultural machinery path conflict information and Dubins path algorithm to determine the obstacle avoidance path for the operation plan that involves detouring around each other;

[0148] Step 210: Calculate the operation path of a single agricultural machine based on the traversal order output in step 206 and the obstacle avoidance path output in step 208, and obtain the final path of each combined agricultural machine.

[0149] In this embodiment, the target farmland is divided into multiple work rows based on agricultural machinery parameters and farmland parameters. These work rows are then assigned to multiple agricultural machines scheduled to work together on the target farmland. Under preset constraints, the optimal combined agricultural machinery operation path is determined based on the transfer costs of each machine. These transfer costs include the entry cost for each machine starting from its designated warehouse and entering its assigned work row, the row-changing cost when working through all assigned work rows, and the departure cost when returning to its designated starting point after completing its work. The optimal combined agricultural machinery operation path corresponds to the path taken by the multiple machines starting from their designated warehouse, passing through all assigned work rows, and returning to their designated starting point. This allows for the comprehensive planning of the optimal combined agricultural machinery operation path by integrating the overall costs generated by the combined operation paths of the machines in the farmland, the transfer paths between work rows, and the specific transfer paths of the machines entering or leaving the farmland. This improves the accuracy of the planned path and enhances the environmental adaptability and safety of the autonomous operation of the agricultural machinery.

[0150] Optionally, such as Figure 7As shown in the figure, this application embodiment also provides a combined agricultural machinery operation path planning device 2000, including a processor 2400 and a memory 2200. The memory 2200 stores a program or instructions that can be run on the processor 2400. When the program or instructions are executed by the processor 2400, they implement the various steps of the above-mentioned combined agricultural machinery operation path planning method embodiment and can achieve the same technical effect. To avoid repetition, it will not be described again here.

[0151] This application also provides a readable storage medium storing a program or instructions. When executed by a processor, the program or instructions implement the various processes of any of the above-described embodiments of the combined agricultural machinery operation path planning method, achieving the same technical effect. To avoid repetition, further details are omitted here. The readable storage medium includes computer-readable storage media, such as read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0152] It should be noted that, in this document, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes that element.

[0153] Through the above description of the embodiments, those skilled in the art can clearly understand that the methods of the above embodiments can be implemented by means of software plus necessary general-purpose hardware platforms. Of course, they can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk) and includes several instructions to cause a terminal (which may be a mobile phone, computer, server, air conditioner, or network device, etc.) to execute the methods described in the various embodiments of this application.

[0154] The embodiments of this application have been described above with reference to the accompanying drawings. However, this application is not limited to the specific embodiments described above. The specific embodiments described above are merely illustrative and not restrictive. Those skilled in the art can make many other forms under the guidance of this application without departing from the spirit and scope of the claims, and all of these forms are within the protection scope of this application.

Claims

1. A method for joint farm machine operation path planning, characterized in that, The method comprises the following steps: dividing a target farmland into a plurality of work rows based on agricultural machine parameters and farmland parameters; allocating the plurality of work rows to a plurality of agricultural machines to be operated in the target farmland; determining an optimal joint agricultural machine operation path based on a transfer cost of each agricultural machine to be operated under a preset constraint condition, wherein the transfer cost comprises an entering cost of each agricultural machine to be operated from a corresponding starting point of a warehouse into a corresponding allocated work row, a row changing cost of each agricultural machine to be operated when passing through all corresponding allocated work rows to perform work, and a leaving cost of each agricultural machine to be operated when returning to the corresponding starting point after completing the work, and the optimal joint agricultural machine operation path corresponds to a path of each agricultural machine to be operated from the corresponding starting point of the warehouse, passing through all corresponding allocated work rows, and returning to the corresponding starting point; wherein the step of allocating the plurality of work rows to the plurality of agricultural machines to be operated in the target farmland comprises: allocating the plurality of work rows to the plurality of agricultural machines to be operated in the target farmland by an outer simulated annealing algorithm of a double-layer simulated annealing algorithm, and specifically comprising: providing the number and position of the plurality of work rows divided by the target farmland and the number of the plurality of agricultural machines to be operated as inputs to the outer simulated annealing algorithm to obtain an outer solution output by the outer simulated annealing algorithm, wherein the outer solution comprises the number and position of the work rows allocated to each agricultural machine to be operated; the step of determining the optimal joint agricultural machine operation path based on the transfer cost of each agricultural machine to be operated comprises: determining a node traversal sequence of each agricultural machine to be operated at a time when the transfer cost is the smallest in the allocated work row by an inner simulated annealing algorithm of the double-layer simulated annealing algorithm to determine the optimal joint agricultural machine operation path, and specifically comprising: passing the outer solution to the inner simulated annealing algorithm, wherein the inner simulated annealing algorithm is used to calculate an inner solution and a corresponding transfer cost at different inner temperatures based on the outer solution; reducing a current inner temperature to a target inner temperature, and adjusting the traversal sequence of the work row allocated to each agricultural machine to be operated in the inner solution to obtain an updated inner solution and a corresponding transfer cost, wherein the inner solution comprises the traversal sequence of the work row allocated to each agricultural machine to be operated corresponding to the outer solution, the target inner temperature comprises one of each inner temperature gradually reduced from an inner initial temperature preset by the inner simulated annealing algorithm and an inner termination temperature preset by the inner simulated annealing algorithm; updating a historically optimal inner solution based on a difference value of the transfer cost corresponding to the inner solution before and after the update, wherein the optimal inner solution corresponds to the smallest transfer cost; if the target inner temperature reaches the inner termination temperature, outputting the historically optimal inner solution, otherwise returning to the step of reducing the current inner temperature to the target inner temperature until the inner termination temperature is reached; and determining the optimal joint agricultural machine operation path based on the traversal sequence of the nodes of the work row corresponding to the output historically optimal inner solution and the number and position of the allocated work rows corresponding to the outer solution.

2. The method of claim 1, wherein, The step of dividing the target farmland into a plurality of work rows based on the agricultural machine parameters and the farmland parameters comprises: selecting a longest boundary in the boundary of the target farmland as a reference line based on farmland parameters of the target farmland; parallel to the reference line, performing equal-interval translation according to a working width of the agricultural machine in the agricultural machine parameters to generate a plurality of working rows until the entire target farmland is covered.

3. The method of claim 1, wherein, The determining, under the preset constraint condition, of the optimal joint agricultural machine working path based on the transfer cost of each to-be-worked agricultural machine includes: taking a starting point of a single to-be-worked agricultural machine in each to-be-worked agricultural machine and a plurality of working rows allocated to the single to-be-worked agricultural machine as nodes; storing a transfer cost between each node in an adjacency matrix, the transfer cost corresponding to an entering cost between a node including the starting point and a node corresponding to a working row, an exiting cost between the node corresponding to the working row and the node including the starting point, and a row-changing cost between the nodes corresponding to the working rows; determining an optimal working path of the single to-be-worked agricultural machine based on a node traversal order when the transfer cost determined by the inner-layer simulated annealing algorithm is the smallest; determining the optimal joint agricultural machine working path based on the optimal working path of each to-be-worked agricultural machine.

4. The method of claim 1, wherein, Before the determining of the optimal joint agricultural machine working path based on the node traversal order of the working row corresponding to the output historical optimal inner-layer solution and the number and position of the allocated working row corresponding to the outer-layer solution, the method further includes: determining the transfer cost corresponding to the outer-layer solution of the outer-layer simulated annealing algorithm based on the transfer cost corresponding to the output historical optimal inner-layer solution and the penalty term corresponding to the outer-layer solution; reducing a current outer-layer temperature to a target outer-layer temperature and adjusting the number and position of the allocated working row for each to-be-worked agricultural machine in the outer-layer solution to obtain an updated outer-layer solution, the target outer-layer temperature including one of each outer-layer temperature gradually reduced from an outer-layer initial temperature preset by the outer-layer simulated annealing algorithm and an outer-layer termination temperature preset by the outer-layer simulated annealing algorithm; returning to the step of passing the outer-layer solution to the inner-layer simulated annealing algorithm based on the updated outer-layer solution until an updated historical optimal inner-layer solution is output by the inner-layer simulated annealing algorithm; determining the transfer cost corresponding to the updated outer-layer solution based on the transfer cost corresponding to the updated historical optimal inner-layer solution and the penalty term corresponding to the outer-layer solution; updating a historical optimal outer-layer solution based on a difference between the transfer costs corresponding to the outer-layer solution before and after the updating, the transfer cost corresponding to the optimal outer-layer solution being the smallest; if the target outer-layer temperature reaches the outer-layer termination temperature, outputting the historical optimal outer-layer solution, otherwise returning to the step of reducing the current outer-layer temperature to the target outer-layer temperature until the outer-layer termination temperature is reached; wherein the determining of the optimal joint agricultural machine working path based on the node traversal order of the working row corresponding to the output historical optimal inner-layer solution and the number and position of the allocated working row corresponding to the outer-layer solution includes: determining the optimal joint agricultural machine working path based on the node traversal order of the working row corresponding to the output historical optimal inner-layer solution and the number and position of the allocated working row corresponding to the output historical optimal outer-layer solution.

5. The method of claim 1, wherein, The method further includes: If there is a conflict between the row-changing paths of two agricultural machines corresponding to work rows in the optimal joint agricultural machine operation path, an obstacle circle with a diameter of the width of an agricultural machine is arranged between two starting points of the two agricultural machines in the respective row-changing paths; Target passing points of the two agricultural machines on the obstacle circle are determined, and the two target passing points have the maximum perpendicular distance to the line connecting the two starting points; Based on the respective starting points and passing points of the two agricultural machines, the corresponding detour paths of the two agricultural machines on the row-changing paths are determined through the Dubins path algorithm.

6. The method of claim 1, wherein, A target agricultural machine to be operated is assigned a plurality of work rows, and a row-changing cost corresponding to the target agricultural machine is related to different turning modes, and the row-changing cost includes one or more combinations of distance cost, time cost, and reserved turning distance generated by the corresponding turning mode.

7. A combined agricultural machine operation path planning device characterized by comprising: A processor and a memory are included, the memory stores programs or instructions executable on the processor, and the programs or instructions are executed by the processor to implement the steps of the method according to any one of claims 1-6.

Images