Regional agricultural machinery resource allocation method based on supply and demand matching

Abstract

The invention discloses a regional agricultural machinery resource allocation method based on supply-demand matching, and belongs to the field of agricultural machinery scheduling. The problems of long scheduling time and high cost in existing cross-regional agricultural machinery resource allocation are solved. According to the method, on the basis of to-be-operated field piece information, agricultural machinery types, initial warehouse position information and termination warehouse position information, the maximum continuous operation duration, cross-regional dynamic deployment, supply and demand matching quality and a farmland demand time window are used as constraints, and the minimum total scheduling cost and total scheduling time are used as targets; constructing a regional agricultural machinery resource allocation model based on supply and demand matching; solving the regional agricultural machinery resource allocation model based on supply and demand matching based on a multi-subswarm particle swarm optimization algorithm to obtain an optimal allocation scheme of agricultural machinery resources; and according to the optimal allocation scheme of the agricultural machinery resources, scheduling the agricultural machinery to start from the initial warehouse, go to the allocated field for operation, and return to the termination warehouse after all allocation tasks are completed. The method is suitable for multi-agricultural-machine operation scheduling.

Description

Technical Field

[0001] This invention belongs to the field of agricultural machinery scheduling. Background Technology

[0002] With the rapid development of smart agriculture and sustainable agriculture, the scale and intensification of agricultural production are constantly increasing. Against this backdrop, the efficient allocation of shared agricultural machinery resources has become a key link in improving agricultural production efficiency. In recent years, intelligent agricultural machinery dispatching platforms based on the Internet of Things (IoT), Global Positioning System (GPS), and cloud computing technologies have been widely used, providing an effective technical approach for cross-regional sharing of agricultural machinery resources. Farmers can submit their farmland operation requests online through such platforms, and the platform's dispatch center is responsible for rationally allocating dispersed agricultural machinery resources to meet the operational needs of various farmlands. Specifically, farmers submit their farmland operation requests online through the platform, and the agricultural machinery dispatch center needs to allocate agricultural machinery resources across regions according to the requests, while meeting the requirements of task time and cost, to complete the operational tasks of these farmlands. However, this method suffers from problems such as long dispatching times and high costs in cross-regional agricultural machinery resource allocation. Summary of the Invention

[0003] This invention aims to solve the problems of long scheduling time and high cost in the existing cross-regional agricultural machinery resource allocation, and proposes a regional agricultural machinery resource allocation method based on supply and demand matching.

[0004] The regional agricultural machinery resource allocation method based on supply and demand matching described in this invention includes:

[0005] Based on information about fields to be operated, types of agricultural machinery, and locations of initial and final warehouses, a regional agricultural machinery resource allocation model based on supply and demand matching is constructed with constraints such as maximum continuous operation time, cross-regional dynamic allocation, supply and demand matching quality, and farmland demand time window, and with the goal of minimizing total scheduling cost and total scheduling time.

[0006] The optimal allocation scheme for agricultural machinery resources is obtained by solving the regional agricultural machinery resource allocation model based on supply and demand matching using the multi-subgroup particle swarm optimization algorithm.

[0007] According to the optimal allocation scheme of agricultural machinery resources, the agricultural machinery is dispatched from the initial warehouse to the assigned field to carry out operations, and returns to the termination warehouse after completing all assigned tasks.

[0008] Furthermore, in this invention, the information of the field to be operated includes the geographical location information of the field to be operated, the area of ​​the field to be operated, the type of operation, the operation parameter requirements, the fixed operation time, and the operation deadline.

[0009] Furthermore, in this invention, the regional agricultural machinery resource allocation model based on supply and demand matching is as follows:

[0010] (1)

[0011] Constraints:

[0012] (2)

[0013] (3)

[0014] (4)

[0015] (5)

[0016] (6)

[0017] (7)

[0018] (8)

[0019] (9)

[0020] (10)

[0021] in, For agricultural machinery collection, For the collection of fields, For a set of nodes, For a set of regions, For the initial garage, To terminate the garage, For agricultural machinery From node arrive Travel costs For agricultural machinery From node arrive Travel time; For agricultural machinery At the node Fixed costs of the operation; For agricultural machinery At the node Fixed time for assignments; For agricultural machinery Maximum continuous working time; For agricultural machinery With fields Supply and demand matching quality indicators; This is the minimum threshold for matching supply and demand. For nodes , Whether it crosses regions; if it does, the value is 1; otherwise, the value is 0. The additional costs and time weighting incurred by cross-regional allocation; The weights of the total scheduling cost in the objective function; The weights of the total scheduling time in the objective function; A variable that is either 0 or 1; when it is 0, it represents agricultural machinery. Not from node arrive A value of 1 indicates agricultural machinery. From node arrive ; For agricultural machinery Complete Node The time point of the task (including the end of the task); i, j, k are node indices; The variable is either 0 or 1. When it is 0, it means that the farm machine m has not traveled from the initial garage O to node j. When it is 1, it means that the farm machine m has traveled from the initial garage O to node j. The variable is either 0 or 1. When it is 0, it means that the farm machine m has not traveled from node i to the terminal garage D. When it is 1, it means that the farm machine m has traveled from node i to the terminal garage D. The variable is either 0 or 1. When it is 0, it means that the agricultural machine m has not traveled from node j to node k. When it is 1, it means that the agricultural machine m has traveled from node j to node k. For agricultural machinery The point in time from the initial garage O; For agricultural machinery The point in time after completing task i at node; Indicates agricultural machinery Travel time from node j to the final garage D; Z represents the task deadline for node j, and Z represents the overall goal. This indicates whether agricultural machinery m is allowed to be transferred across regions.

[0022] Furthermore, in this invention, the method for solving the regional agricultural machinery resource allocation model based on supply and demand matching using a multi-subgroup particle swarm optimization algorithm to obtain the optimal allocation scheme for agricultural machinery resources is as follows:

[0023] Step 1: Initialize the particle swarm, where each particle represents an agricultural machinery scheduling scheme, which includes agricultural machinery task allocation and path sequence information;

[0024] The particle position is defined to represent the current scheduling scheme, and the particle velocity represents the direction and magnitude of the change in the scheduling scheme.

[0025] Step 2: Divide the particle swarm evenly into multiple subgroups. Based on the regional agricultural machinery resource allocation model that matches supply and demand, assign differentiated optimization objectives to each subgroup. Based on the objectives, total scheduling cost, and total scheduling time of each subgroup, combined with the penalty term for constraint violation, establish a fitness function for each subgroup.

[0026] Step 3: Each subgroup obtains the fitness value of each particle within the subgroup according to its own fitness function, and updates the velocity of each particle using the fitness value;

[0027] Step 4: Update the position of each particle based on its current velocity. After each update iteration, collect the individual values ​​corresponding to the optimal positions of all particles in all subgroups. Furthermore, every T generations, individuals in the subgroup at the optimal position are exchanged through a migration operation. Return to step three and continue until the iteration termination condition is met. Select the particle with the highest fitness value in all subgroups as the optimal particle and obtain the optimal allocation scheme.

[0028] This invention integrates four major constraints, including maximum continuous operation time and cross-regional allocation, effectively reducing total cost and total working time while meeting these constraints. Simultaneously, this invention employs a multi-subgroup particle swarm optimization algorithm for global optimization, ensuring the quality of the allocation plan, reducing the time spent on manual scheduling, and effectively improving the efficiency of agricultural machinery operations. Attached Figure Description

[0029] Figure 1 This is a flowchart of a regional agricultural machinery resource allocation method based on supply and demand matching as described in this invention. Detailed Implementation

[0030] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention. It should be noted that, unless otherwise specified, the embodiments and features in the embodiments of the present invention can be combined with each other.

[0031] Specific implementation method one: Refer to Figure 1 This embodiment specifically describes a regional agricultural machinery resource allocation method based on supply and demand matching, which includes:

[0032] Based on information about fields to be operated, types of agricultural machinery, and locations of initial and final warehouses, a regional agricultural machinery resource allocation model based on supply and demand matching is constructed with constraints such as maximum continuous operation time, cross-regional dynamic allocation, supply and demand matching quality, and farmland demand time window, and with the goal of minimizing total scheduling cost and total scheduling time.

[0033] The optimal allocation scheme for agricultural machinery resources is obtained by solving the regional agricultural machinery resource allocation model based on supply and demand matching using the multi-subgroup particle swarm optimization algorithm.

[0034] According to the optimal allocation scheme of agricultural machinery resources, the agricultural machinery is dispatched from the initial warehouse to the assigned field to carry out operations, and returns to the termination warehouse after completing all assigned tasks.

[0035] Furthermore, in this invention, the information of the field to be operated includes the geographical location information of the field to be operated, the area of ​​the field to be operated, the type of operation, the operation parameter requirements, the fixed operation time, and the operation deadline.

[0036] Furthermore, in this invention, the regional agricultural machinery resource allocation model based on supply and demand matching is as follows:

[0037] (1)

[0038] Constraints:

[0039] (2)

[0040] (3)

[0041] (4)

[0042] (5)

[0043] (6)

[0044] (7)

[0045] (8)

[0046] (9)

[0047] (10)

[0048] in, For agricultural machinery collection, For the collection of fields, For a set of nodes, For a set of regions, For the initial garage, To terminate the garage, For agricultural machinery From node arrive Travel costs For agricultural machinery From node arrive Travel time; For agricultural machinery At the node Fixed costs of the operation; For agricultural machinery At the node Fixed time for assignments; For agricultural machinery Maximum continuous working time; For agricultural machinery With fields Supply and demand matching quality indicators; This is the minimum threshold for matching supply and demand. For nodes , Whether it crosses regions; if it does, the value is 1; otherwise, the value is 0. The additional costs and time weighting incurred by cross-regional allocation; The weights of the total scheduling cost in the objective function; The weights of the total scheduling time in the objective function; A variable that is either 0 or 1; when it is 0, it represents agricultural machinery. Not from node arrive A value of 1 indicates agricultural machinery. From node arrive ; For agricultural machinery Complete Node The time point of the task (including the end of the task); i, j, k are node indices; The variable is either 0 or 1. When it is 0, it means that the farm machine m has not traveled from the initial garage O to node j. When it is 1, it means that the farm machine m has traveled from the initial garage O to node j. The variable is either 0 or 1. When it is 0, it means that the farm machine m has not traveled from node i to the terminal garage D. When it is 1, it means that the farm machine m has traveled from node i to the terminal garage D. The variable is either 0 or 1. When it is 0, it means that the agricultural machine m has not traveled from node j to node k. When it is 1, it means that the agricultural machine m has traveled from node j to node k. For agricultural machinery The point in time from the initial garage O; For agricultural machinery The point in time after completing task i at node; Indicates agricultural machinery Travel time from node j to the final garage D; Z represents the task deadline for node j, and Z represents the overall goal. This indicates whether agricultural machinery m is allowed to be transferred across regions.

[0049] Furthermore, in this invention, the method for solving the regional agricultural machinery resource allocation model based on supply and demand matching using a multi-subgroup particle swarm optimization algorithm to obtain the optimal allocation scheme for agricultural machinery resources is as follows:

[0050] Step 1: Initialize the particle swarm, where each particle represents an agricultural machinery scheduling scheme, which includes agricultural machinery task allocation and path sequence information;

[0051] The particle position is defined to represent the current scheduling scheme, and the particle velocity represents the direction and magnitude of the change in the scheduling scheme.

[0052] Step 2: Divide the particle swarm evenly into multiple subgroups. Based on the regional agricultural machinery resource allocation model that matches supply and demand, assign differentiated optimization objectives to each subgroup. Based on the objectives, total scheduling cost, and total scheduling time of each subgroup, combined with the penalty term for constraint violation, establish a fitness function for each subgroup.

[0053] Step 3: Each subgroup obtains the fitness value of each particle within the subgroup according to its own fitness function, and updates the velocity of each particle using the fitness value;

[0054] Step 4: Update the position of each particle based on its current velocity. After each update iteration, collect the individual values ​​corresponding to the optimal positions of all particles in all subgroups. Furthermore, every T generations, individuals in the subgroup at the optimal position are exchanged through a migration operation. Return to step three and continue until the iteration termination condition is met. Select the particle with the highest fitness value in all subgroups as the optimal particle and obtain the optimal allocation scheme.

[0055] In actual agricultural production, the efficient operation of agricultural machinery is crucial to ensuring agricultural productivity. However, in practice, the cross-regional allocation of agricultural machinery presents problems such as mismatch between supply and demand, high operating costs, and long operating times. To address these issues, cross-regional allocation of agricultural machinery resources, based on consideration of supply and demand matching, can more rationally utilize these resources and ensure the efficient completion of agricultural production activities.

[0056] This scheduling problem involves multiple heterogeneous agricultural machines and multiple farmlands. The goal is to minimize the total working time and total operating cost of the agricultural machines while satisfying constraints such as maximum continuous working time, cross-regional dynamic matching, supply and demand matching quality, and farmland demand time window.

[0057] This problem has the following characteristics:

[0058] 1. Maximum continuous operating time constraint: Considering that agricultural machinery cannot operate continuously in actual agricultural production, the multiple heterogeneous agricultural machines involved in this invention need to perform tasks such as refueling and maintenance during operation. Each agricultural machine has a different maximum continuous operating time, and the total working time of each agricultural machine cannot exceed the maximum continuous operating time.

[0059] 2. Cross-regional dynamic matching constraints: Each agricultural machine has a limited working area. If cross-regional operations are carried out, the required operation time will be longer and the operation cost will be higher.

[0060] 3. Supply-Demand Matching Quality Constraint: Considering that agricultural production activities may differ across fields, and different agricultural machines may be capable of performing different agricultural activities, this model assumes that the operational capacity of each agricultural machine and the operational needs of each field have different degrees of matching, i.e., a supply-demand matching index. Only agricultural machines that meet the minimum supply-demand matching quality index can operate in the target field.

[0061] 4. Farmland Demand Time Window: Considering that different fields have different operation priorities in actual production, this model assumes that each field has an operation time window, and agricultural machinery needs to complete the operation of the target field within the specified latest working time.

[0062] Construct a corresponding model based on the characteristics of the above problems;

[0063] 1. Problem Description: Each agricultural machine m starts from the initial warehouse O, performs operations between various nodes, and returns to the termination warehouse D after completing the requirements of all fields while satisfying all constraints.

[0064] Agricultural machinery set M: Agricultural machinery m is used to complete the work of each field and has the following characteristics: (1) Each agricultural machinery has different working capabilities, and moving at each node will generate different fixed moving costs and fixed moving time.

[0065] (2) Each agricultural machine has its own working area. When operating on its own area or across areas, it will generate different fixed operating time and fixed operating costs. If it operates across areas, the required operating time will be longer and the operating cost will be higher.

[0066] (3) Each agricultural machine has a different maximum continuous working time, and the total working time of each agricultural machine shall not exceed the maximum continuous working time.

[0067] (4) Each agricultural machine can only go to the field that meets the minimum supply and demand matching quality index to carry out operations.

[0068] Field set F: Field f represents the task point with operation requirements, and has the following characteristics: (1) Each field has a latest task deadline, and the agricultural machinery must arrive and complete the operation before this deadline.

[0069] (2) Each field can only be visited once, and all requirements must be met.

[0070] (3) Considering that in actual production, the size of the fields varies, and each agricultural machine has different fixed operating costs and fixed operating time when operating in any field.

[0071] 2. Model Construction: (1) Minimum total scheduling cost. The total scheduling cost is the sum of the total travel cost and the total operation cost. (2) Minimum total scheduling time. The total scheduling time is the sum of the total travel time and the total operation time.

[0072] For the regional agricultural machinery resource allocation problem based on supply and demand matching, the core idea of ​​this algorithm is to represent a complete set of scheduling schemes as a particle, and then move towards a better solution by updating the position and velocity, and finally find the optimal scheduling scheme.

[0073] 3. Parameter Assumptions: Indicates a collection of agricultural machinery; Represents a set of fields; Represents a set of nodes; Represents a set of regions; Indicates the initial garage; Indicates termination of garage service; Indicates agricultural machinery From node arrive Travel costs; Indicates agricultural machinery From node arrive Travel time; Indicates agricultural machinery At the node Fixed costs of the operation; Indicates agricultural machinery At the node Fixed time for assignments; Indicates agricultural machinery Maximum continuous working time; Represents a node Priority weights; Indicates agricultural machinery With fields Supply and demand matching quality indicators; This represents the minimum threshold for matching supply and demand. Represents a node , Whether it crosses regions; if it does, the value is 1; otherwise, the value is 0. Indicates the additional costs and time weighting resulting from cross-regional allocation; The weights representing the total scheduling cost in the objective function; This represents the weight of the total scheduling time in the objective function; Representing 0-1 variables, agricultural machinery From node arrive ; Indicates agricultural machinery Complete Node The timeframe for completing the assignment (including the completion date); Represents an infinitely large variable;

[0074] 4. The expression for the objective function is:

[0075] (1)

[0076] The objective function (1) is the overall scheduling objective, namely, minimizing the total cost and the shortest total operation time.

[0077] 5. Constraints:

[0078] (2)

[0079] (3)

[0080] (4)

[0081] (5)

[0082] (6)

[0083] (7)

[0084] (8)

[0085] (9)

[0086] (10)

[0087] Constraint (2) is a path closure constraint, which stipulates that each agricultural machine starts from the initial garage, completes the assigned task, and returns to the termination garage; Constraint (3) is a node balance flow constraint, which stipulates that the number of inflows and outflows of each node is one; Constraint (4) stipulates that the continuous working time of each agricultural machine is 0 when it starts from the initial garage; Constraint (5) is an increasing constraint on the continuous working time of agricultural machines; Constraint (6) stipulates that the continuous working time cannot exceed the longest continuous working time; Constraint (7) is a time sensitivity and priority constraint, which means that the time when the agricultural machine completes each field cannot be later than the task deadline; Constraint (8) is a supply and demand matching quality constraint, which stipulates that only when the matching quality meets the standard can the agricultural machine go to the field to work; Constraint (9) is a cross-regional dynamic allocation constraint, which determines whether the agricultural machine is allowed to be allocated across regions; Constraint (10) stipulates the range of values ​​for each variable.

[0088] The model was solved using multi-subgroup particle swarm optimization.

[0089] Multi-subgroup particle swarm optimization (MSO) is suitable for complex, multi-constraint, and discontinuous scheduling problems. It simulates the foraging behavior of flocks of birds or schools of fish, and gradually approaches the optimal scheduling solution through the cooperative search of particle swarms in the solution space.

[0090] 1. Particle: Represents a complete scheduling scheme, including the task allocation and path order of each agricultural machine.

[0091] 2. Position: The current particle's solution, i.e., the current scheduling and path arrangement of agricultural machinery.

[0092] 3. Velocity: The direction and magnitude of the solution change, used to guide how particles adjust the scheduling scheme (such as changing paths, changing vehicles, changing order).

[0093] 4. Individual Optimal: The particle's own historical optimal solution.

[0094] 5. Global optimum: the best historical solution of the particle swarm.

[0095] 6. Subgroups: The particle swarm is divided into several subgroups, each focusing on a different optimization objective (such as optimizing the path, optimizing the number of vehicles, optimizing cross-regional allocation, etc.).

[0096] Step 1: Particle encoding and initialization;

[0097] In particle swarm optimization (PSO), the particle position P is encoded as a discrete sequence containing the node visit order, representing the allocation of farmland node F to two farm machines. During initialization, a random feasible node sequence is generated for each particle, ensuring that each farmland node j∈F appears exactly once, and that each path begins with O and ends with D. In multi-subgroup particle swarm optimization (MPSO), the particle swarm is divided into several subgroups, each focusing on optimizing a different direction. For example, one subgroup might focus on reducing the number of farm machines, while another focuses on optimizing the node order, thus increasing search diversity.

[0098] Step 2: Calculate the objective function and check constraints;

[0099] The fitness function is used to evaluate the quality of the scheduling solution for each particle. First, the particle needs to be decoded, which involves splitting the node sequence according to OD (Original Distributed Origin) to obtain the access paths for each machine m ∈ M. Then, based on these paths, the total travel cost, total operating cost, total travel time, and total operating time for each machine are calculated. The fitness value is usually set to the negative of the objective function (because PSO defaults to maximizing fitness), therefore, for cost minimization problems, the fitness value is negative. If a particle solution violates the maximum operating time or node deadline, a large penalty term is added to the fitness function, or the solution is directly repaired during the decoding process, such as adjusting the node order or splitting paths, to ensure the feasibility of the solution.

[0100] Step 3: Particle velocity and position update;

[0101] To accommodate the discrete nature of the vehicle path problem, the particle's velocity *v* is defined as a series of discrete operations, rather than directly using the addition formula in continuous space. In each iteration, the particle position... Based on the current speed Updated to ,in This indicates the application of the operation sequence. The influence of individual and group extrema on velocity in the particle swarm is transformed into generating a sequence of difference operations for the current solution, thereby guiding the solution to move towards a better direction in the discrete space.

[0102] Step 4: Core Improvements;

[0103] In the multi-subgroup cooperation mechanism, the population S is divided into s subgroups. Each subgroup independently optimizes a specific objective, and every T generations, excellent individuals from each subgroup are exchanged through a migration operation. To enhance information sharing and global search capabilities, MPSO integrates a hybrid local search mechanism to refine and optimize the current global optimum after each iteration. To dynamically balance exploration and development, the algorithm adaptively adjusts the learning factor based on population diversity. (Individual experience weight) and (Group experience weight), increases as diversity decreases. To promote individual exploration and reduce To reduce the pressure of group convergence.

[0104] Step 5: Constraint handling and repair;

[0105] Constraint handling is implemented throughout the entire decoding and update process. When decoding particle P, a path segmentation algorithm is first used to accumulate the working time and travel time of each node j in the path of each agricultural machine from left to right. Once the maximum working time is exceeded, the path is forcibly split by inserting the termination node D and the starting node O to form a new train number. For time window constraints, if the current order causes a violation of the deadline, it is corrected first by local adjustment (such as swapping the order of adjacent nodes), and a certain penalty is imposed if necessary. Cross-regional allocation constraints require that the number of times agricultural machines visit across regions does not exceed a preset threshold; otherwise, the task must be reassigned to agricultural machines within the current region.

[0106] Step 6: Termination Condition;

[0107] The algorithm terminates after reaching the maximum number of iterations.

[0108] Step 7: Output the optimal solution;

[0109] The algorithm ultimately outputs the agricultural machinery scheduling scheme encoded by the globally optimal particle, which specifically includes the path of each agricultural machine m, the cumulative travel time of the path, the operation time, the travel cost, the operation cost, as well as the overall total scheduling cost and total scheduling time.

[0110] This invention employs a multi-subgroup particle swarm optimization algorithm, significantly improving the global optimal solution acquisition rate. Compared to the traditional single-subgroup PSO algorithm's tendency to get trapped in local optima, this invention expands the search coverage through multi-subgroup differentiated exploration (cost, time, and constraint-compliant subgroup division of labor) + a cross-subgroup excellent individual migration mechanism. This effectively avoids the suboptimal scheduling scheme problem caused by premature convergence of the algorithm. Furthermore, the faster iterative convergence speed shortens the scheduling scheme generation time, meeting the rapid decision-making needs during busy farming seasons. It achieves the scheduling goals of reducing agricultural machinery operating costs and minimizing operation time. Through multi-subgroup collaborative optimization, the output scheduling scheme can achieve a "cost-time" balance optimization.

[0111] While the invention has been described herein with reference to specific embodiments, it should be understood that these embodiments are merely examples of the principles and applications of the invention. Therefore, it should be understood that many modifications can be made to the exemplary embodiments, and other arrangements can be designed without departing from the spirit and scope of the invention as defined by the appended claims. It should be understood that different dependent claims and features described herein can be combined in ways different from those described in the original claims. It is also understood that features described in conjunction with individual embodiments can be used in other described embodiments.

Claims

1. A method for regional agricultural machinery resource allocation based on supply and demand matching, characterized in that, The method includes: Based on information about fields to be operated, types of agricultural machinery, and locations of initial and final warehouses, a regional agricultural machinery resource allocation model based on supply and demand matching is constructed with constraints such as maximum continuous operation time, cross-regional dynamic allocation, supply and demand matching quality, and farmland demand time window, and with the goal of minimizing total scheduling cost and total scheduling time. The optimal allocation scheme for agricultural machinery resources is obtained by solving the regional agricultural machinery resource allocation model based on supply and demand matching using the multi-subgroup particle swarm optimization algorithm. According to the optimal allocation scheme of agricultural machinery resources, the agricultural machinery is dispatched from the initial warehouse to the assigned field to carry out operations, and returns to the termination warehouse after completing all assigned tasks.

2. The method for regional agricultural machinery resource allocation based on supply and demand matching according to claim 1, characterized in that... The information on fields to be operated includes the geographical location of the fields to be operated, the area of ​​the fields to be operated, the type of operation, the required operation parameters, the fixed operation time, and the deadline for operation.

3. The method for regional agricultural machinery resource allocation based on supply and demand matching according to claim 1, characterized in that, The regional agricultural machinery resource allocation model based on supply and demand matching is as follows: （1） Constraints: （2） （3） （4） （5） （6） （7） （8） （9） （10） in, For agricultural machinery collection, For the collection of fields, For a set of nodes, For a set of regions, For the initial garage, To terminate the garage, For agricultural machinery From node arrive Travel costs For agricultural machinery From node arrive Travel time; For agricultural machinery At the node Fixed costs of the operation; For agricultural machinery At the node Fixed time for assignments; For agricultural machinery Maximum continuous working time; For agricultural machinery With fields Supply and demand matching quality indicators; This is the minimum threshold for matching supply and demand. For nodes , Whether it crosses regions; if it does, the value is 1; otherwise, the value is 0. The additional costs and time weighting incurred by cross-regional allocation; The weights of the total scheduling cost in the objective function; The weights of the total scheduling time in the objective function; A variable that is either 0 or 1; when it is 0, it represents agricultural machinery. Not from node arrive A value of 1 indicates agricultural machinery. From node arrive ; For agricultural machinery Complete Node The time point of the task (including the end of the task); i, j, k are node indices; The variable is either 0 or 1. When it is 0, it means that the farm machine m has not traveled from the initial garage O to node j. When it is 1, it means that the farm machine m has traveled from the initial garage O to node j. The variable is either 0 or 1. When it is 0, it means that the farm machine m has not traveled from node i to the terminal garage D. When it is 1, it means that the farm machine m has traveled from node i to the terminal garage D. The variable is either 0 or 1. When it is 0, it means that the agricultural machine m has not traveled from node j to node k. When it is 1, it means that the agricultural machine m has traveled from node j to node k. For agricultural machinery The point in time from the initial garage O; For agricultural machinery The point in time after completing task i at node; Indicates agricultural machinery Travel time from node j to the final garage D; Z represents the task deadline for node j, and Z represents the overall goal. This indicates whether agricultural machinery m is allowed to be transferred across regions.

4. The method for regional agricultural machinery resource allocation based on supply and demand matching according to claim 1, characterized in that, The method for solving the regional agricultural machinery resource allocation model based on supply and demand matching using the multi-subgroup particle swarm optimization algorithm to obtain the optimal allocation scheme for agricultural machinery resources is as follows: Step 1: Initialize the particle swarm, where each particle represents an agricultural machinery scheduling scheme, which includes agricultural machinery task allocation and path sequence information; The particle position is defined to represent the current scheduling scheme, and the particle velocity represents the direction and magnitude of the change in the scheduling scheme. Step 2: Divide the particle swarm evenly into multiple subgroups. Based on the regional agricultural machinery resource allocation model that matches supply and demand, assign differentiated optimization objectives to each subgroup. Based on the objectives, total scheduling cost, and total scheduling time of each subgroup, combined with the penalty term for constraint violation, establish a fitness function for each subgroup. Step 3: Each subgroup obtains the fitness value of each particle within the subgroup according to its own fitness function, and updates the velocity of each particle using the fitness value; Step 4: Update the position of each particle based on its current velocity. After each update iteration, collect the individual values ​​corresponding to the optimal positions of all particles in all subgroups. Furthermore, every T generations, individuals in the subgroup at the optimal position are exchanged through a migration operation. Return to step three and continue until the iteration termination condition is met. Select the particle with the highest fitness value in all subgroups as the optimal particle and obtain the optimal allocation scheme.

Images