Ridger scheduling method for soil mechanical compaction reduction

Abstract

The application relates to the technical field of agricultural machinery scheduling, and discloses a ridger scheduling method for soil mechanical compaction reduction, which is used for solving the soil compaction problem caused by unreasonable ridger scheduling in the crop planting stage in the prior art and repeatedly rolling in the field. In the technical scheme, a scheduling model is solved with the minimum soil compaction degree and operation time as the target, so that the optimal ridger scheduling scheme is sought, and the soil compaction degree is reduced. In the technical scheme, field block information, warehouse information and ridger information are taken as inputs, the scheduling model is solved through a preset algorithm with the minimum soil compaction degree and operation time as the target, so that the ridger scheduling scheme is obtained, the operation time of the ridger and the soil compaction are fully considered, the ridger is scheduled with the least non-operation distance, and the soil compaction damage is reduced.

Description

Technical Field

[0001] This invention relates to the field of agricultural machinery scheduling technology, and in particular to a method for scheduling ridging machines to reduce soil compaction. Background Technology

[0002] Soil is the foundation of agricultural production, and its healthy physical structure directly affects crop growth and development, water retention, and nutrient cycling. However, in modern intensive agricultural production, the frequent use of large agricultural machinery, while increasing efficiency, has also led to serious soil compaction problems. Repeated compaction by heavy equipment in the field damages soil aggregate structure, resulting in increased soil bulk density, reduced porosity, and poor aeration and permeability, ultimately forming a hard plow pan. This severely restricts crop root development and water and fertilizer absorption, becoming one of the key obstacles affecting the sustainable development of global agriculture.

[0003] Ridging is a crucial step in the planting of many crops (such as potatoes, tobacco, corn, and vegetables), aiming to improve the soil microenvironment, increase soil temperature, facilitate drainage, and aid in field management. However, in current technologies, particularly in smart agriculture, the timing of ridge-making operations and the varying locations of storage warehouses and fields waiting to be worked can lead to repeated compaction of the equipment in the field, especially in the same area, such as furrows. This repeated compaction exacerbates soil degradation.

[0004] Therefore, how to achieve more reasonable scheduling of ridging machines and reduce soil compaction damage is a problem that urgently needs to be solved by those skilled in the art. Summary of the Invention

[0005] The purpose of this invention is to solve the problem of soil compaction caused by repeated rolling in the field due to unreasonable scheduling of ridging machines during the crop planting stage in the current technology. Therefore, this invention provides a method for scheduling ridging machines to reduce soil mechanical compaction, which aims to minimize soil compaction and the total operation time, thereby reducing soil compaction damage.

[0006] To address the aforementioned technical problems, this invention provides a method for scheduling ridging machines to reduce soil mechanical compaction, comprising:

[0007] Obtain information on field plots, warehouses, and ridging machines;

[0008] A preset algorithm is invoked to solve the scheduling model based on the field information, warehouse information, and ridging machine information, with the goal of minimizing soil compaction and operation time, to obtain a ridging machine scheduling scheme.

[0009] Preferably, the objective function of the scheduling model is:

[0010] ;

[0011] in, ;

[0012] ;

[0013] For the comprehensive cost function, This represents the total soil compaction degree. Total operation time Weights based on decision-maker preferences. It is a time constant. To improve compaction sensitivity, The set of coordinates for all fields. This is the average of the total number of times all points within the field operation area are rolled by the tires of all ridging machines; For ridging machine assembly, For ridging machine The time to start the assignment. For ridging machine The time to finish the task.

[0014] Preferably, the constraints of the scheduling model include:

[0015] ;

[0016] ;

[0017] ;

[0018] ;

[0019] ;

[0020] ;

[0021] in, Indicates a ridging machine In the field Whether or not to perform the task is a decision variable; For the collection of fields; Indicates allocation to the ridging machine In the field The number of furrows on which the work needs to be completed; For ridging machine The working width; For fields The width; For the indicator function, if point Located in the ridging machine From node arrive If the value is 1 on the path segment, it is 0 otherwise. The nodes include field entrance point, field exit point and warehouse point. For ridging machine Reaching the node Time; For ridging machine Leave node Time; Represents a node and The distance the ridging machine travels between them; Indicates a ridging machine From node Move to The speed of travel; For ridging machine The starting point of the workable time window, For ridging machine Time of leaving the warehouse For ridging machine The time to return to the warehouse For ridging machine The end of the workable time window; For ridging machine Total operation time For ridging machine Unit time operation cost For the total budget.

[0022] Preferably, the preset algorithm is: PPO algorithm combined with simulated annealing algorithm;

[0023] The invocation of the preset algorithm includes: conducting a global exploration using the PPO algorithm to make decisions on the allocation of tasks and path selection for the ridging machine; in the action selection stage of the PPO algorithm, using the simulated annealing algorithm to perturb the initial action to generate a new solution and calculate the cost function, and determining whether to accept the new solution based on the cost function.

[0024] Preferably, the formula for calculating the cost function is:

[0025] ;

[0026] Indicates the action state The sum of time cost and soil compaction cost, This is the time cost weighting coefficient. To solidify the cost weighting coefficient, Action state Total homework time Action state The total soil compaction degree.

[0027] Preferably, the step of determining whether to accept the new solution based on the cost function includes:

[0028] Calculate the difference in cost functions between the two actions: ;

[0029] like Then always accept ,like Then, based on probability accept ;

[0030] in, The new solution represents the action state. The sum of time cost and soil compaction cost, To simulate the current temperature parameters of the annealing algorithm, It is a natural constant.

[0031] Preferably, in the interaction step between the PPO algorithm and the environment, the formula for calculating the cost reward is:

[0032] ;

[0033] As a cost incentive; This refers to the time-based reward weighting coefficient. As an efficiency bonus, the value is a negative number representing the increase in total task time; To solidify the reward weighting coefficient; As an ecological reward, the value is a negative number representing the increase in total soil compaction. To constrain penalties, if the action Violation of any hard constraints It is a preset maximum negative reward, otherwise it is 0.

[0034] Preferably, the hard constraints include: total budget overruns and workable time window conflicts.

[0035] Preferred type: ridging machine In two consecutive fields and The time constraint for the transition between them is:

[0036] ;

[0037] in, For ridging machine Leave the field Time at the exit point For ridging machine Arrival at the field Entry point time, For fields The coordinates of the exit point, For fields The coordinates of the entry point, For ridging machine The empty transfer speed.

[0038] Preferably, the warehouse information includes: the number of warehouses, warehouse coordinates, and the maximum number of ridging machines that can be allocated to each warehouse;

[0039] The field information includes: number of fields, field length, field width, minimum number of ridging strips per field, field entrance point, and field exit point;

[0040] The information on the ridging machine includes: the number of ridging machines, the working width of the ridging machine, the field working speed, the empty transfer speed, the operating cost per unit time, and the available working time window.

[0041] This invention provides a method for scheduling ridging machines to reduce soil compaction. Compared to current technologies where improper scheduling of ridging machines during the crop planting stage leads to repeated compaction in the field, this method aims to minimize soil compaction and operation time by solving a scheduling model to find the optimal ridging machine scheduling scheme, thereby reducing soil compaction. Using this method, field information, warehouse information, and ridging machine information are taken as input. A preset algorithm is used to solve the scheduling model with the goal of minimizing soil compaction and operation time, resulting in a ridging machine scheduling scheme. This scheme fully considers the duration of ridging machine operation and soil compaction, achieving ridging machine scheduling with minimal non-operational distance, thus reducing soil compaction damage. Attached Figure Description

[0042] To more clearly illustrate the embodiments of this application, the accompanying drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0043] Figure 1 A flowchart of a ridging machine scheduling method for reducing soil mechanical compaction, provided as an embodiment of the present invention. Detailed Implementation

[0044] 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 of ordinary skill in the art without creative effort are within the protection scope of the present invention.

[0045] The core of this invention is to provide a method for scheduling ridging machines to reduce soil mechanical compaction. The method aims to minimize soil compaction and reduce total operating time, thereby reducing soil compaction damage.

[0046] To enable those skilled in the art to better understand the present invention, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0047] Figure 1 A flowchart of a ridging machine scheduling method for reducing soil mechanical compaction provided by an embodiment of the present invention is shown below. Figure 1 As shown, the method includes:

[0048] S10: Obtain field information, warehouse information, and ridging machine information;

[0049] S11: Call the preset algorithm to solve the scheduling model based on field information, warehouse information and ridging machine information, with the goal of minimizing soil compaction and operation time, to obtain the ridging machine scheduling scheme.

[0050] The ridging machine scheduling method for reducing soil compaction provided in this application is mainly used during the ridging operation period in crop planting. Through a scheduling model, it seeks a suitable ridging machine scheduling scheme to avoid repeated compaction of the field while ensuring operational needs are met, thereby reducing soil compaction. The execution entity of this method can be a ridging machine scheduling device for reducing soil compaction. This device may specifically include a memory and a processor. The memory stores a computer program, and the processor executes the computer program to implement the steps of the ridging machine scheduling method for reducing soil compaction provided in this application. In some embodiments, the ridging machine scheduling device for reducing soil compaction may also include a display, touchscreen, or other human-computer interaction device. In specific implementations, the ridging machine scheduling device for reducing soil compaction provided in this embodiment may include, but is not limited to, smartphones, tablets, laptops, or desktop computers.

[0051] Of course, it is understood that if the methods in the embodiments of this application are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, all or part of the technical solutions of this application can be embodied in the form of a software product, which is stored in a storage medium and executes all or part of the steps of the methods described in the various embodiments of this application.

[0052] In the technical solution provided in this application, the generation of the ridging machine scheduling scheme relies on a pre-built scheduling model. Constraints are set through factors such as operational requirements and conditions, and the scheduling model is solved with the goal of minimizing soil compaction and operation time, thus obtaining the most suitable scheduling scheme. Therefore, in practical applications, it is first necessary to obtain field information, warehouse information, and ridging machine information. Field information mainly includes the number of fields, field length, field width, minimum number of ridging strips per field, field entry point, and field exit point. Warehouse information mainly includes the number of warehouses, warehouse coordinates, and the maximum number of ridging machines that can be allocated to each warehouse. Ridging machine information mainly includes the number of ridging machines, ridging machine operating width, field operating speed, empty-run transfer speed, unit time operation cost, and available operating time window. Based on the obtained information, a preset algorithm is used to solve the scheduling model with the goal of minimizing soil compaction and operation time to obtain the ridging machine scheduling scheme. In this application, to better calculate and represent soil compaction degree based on the state and form of the ridging machine during operation, the soil compaction degree is determined according to the number of times the field is compacted. The scheduling model in this application is constructed based on the integer proportional relationship between the working width of the ridging machine and the field size, and the soil compaction degree is a value determined according to the number of times the field is compacted.

[0053] The pre-defined algorithm in this application is key to achieving a fast and accurate optimal ridging machine scheduling scheme. This application provides a specific pre-defined algorithm: a combination of Proximal Policy Optimization (PPO) and simulated annealing. Calling the pre-defined algorithm includes: performing global exploration using the PPO algorithm to make decisions on ridging machine task allocation and path selection; in the action selection phase of the PPO algorithm, using simulated annealing to perturb the initial actions to generate new solutions and calculate the cost function, determining whether to accept the new solution based on the cost function. The pre-defined algorithm used in this application fully utilizes PPO's strength in fine-grained policy optimization. Simulated annealing helps the algorithm escape local optima and enhances its global exploration capabilities. PPO, as the main framework, is responsible for global exploration, and the learned policy can intelligently make decisions on task allocation and path selection. The simulated annealing algorithm, as a local search embedded in the action selection phase of PPO, performs local fine-grained optimization on the actions proposed by PPO to improve solution quality and accelerate convergence. When PPO training gets stuck in a local optimum, the "Metropolis criterion" of simulated annealing is used to slightly perturb the policy parameters to attempt to escape the local optimum. This leads to the identification of the optimal ridging machine scheduling scheme that meets the objectives.

[0054] The ridging machine scheduling method provided in this application addresses the soil compaction reduction issue caused by unreasonable ridging machine scheduling during the crop planting stage, which leads to repeated compaction. This method aims to minimize soil compaction and operation time by solving a scheduling model to find the optimal ridging machine scheduling scheme, thereby reducing soil compaction. Using this scheme, field information, warehouse information, and ridging machine information are taken as input. A preset algorithm is used to solve the scheduling model with the goal of minimizing soil compaction and operation time, resulting in a ridging machine scheduling scheme. This scheme fully considers the duration of ridging machine operation and soil compaction, achieving ridging machine scheduling with minimal non-operational distance, thus reducing soil compaction damage.

[0055] The above embodiments introduced the generation of a ridging machine scheduling scheme through a scheduling model. This embodiment provides a specific method for constructing the scheduling model. It is understood that, considering actual ridging machine scheduling needs and minimizing compaction of the field, the working width of the ridging machine should be an integer ratio to the width of the field to avoid the need for repeated, multiple round trips by the ridging machine. In this embodiment, the construction of the scheduling model includes multiple stages. Stage one defines the initialization parameters, including the warehouse set parameters: This indicates the assembly point where the ridging machines are parked in the warehouse. The total number of warehouses. For warehouse Geographical coordinates , Represents warehouse Maximum number of ridging machines that can be accommodated or allocated; plot set parameters: This represents the collection of fields to be processed. The total number of fields, Indicates field width, , Indicates field Length, Indicates field The minimum number of ridgings required to complete is [number]. This indicates that the ridging machine has entered the field. The starting position of the entrance, This indicates that the ridging machine has completed its operation and left the field. Export location; Ridging machine assembly parameters: This represents the set of available ridging machines. This represents the total number of ridging machines. For ridging machine The working width, , Indicates a ridging machine The speed of field operations, Indicates a ridging machine empty transfer speed, Indicates a ridging machine The unit time operation cost, Indicates a ridging machine The workable time window. Phase two is responsible for designing decision variables, including task allocation variables, ridge operation integer variables, path sequence variables, time variables, and soil compaction degree variables; when designing task allocation variables, This indicates that if the variable is 1, then the field... The task was assigned to the ridging machine. Otherwise, it is 0, indicating a ridging machine. Unoperated fields ; A variable of 1 indicates a ridging machine. Assigned from warehouse Depart and eventually return to the warehouse. If the variable is 0, it indicates a ridging machine. With warehouse There is no assignment relationship. When designing integer variables for row operations, Indicates allocation to the ridging machine In the field The number of furrows that need to be completed, if ,but It must be 0, if ,but The field width and ridging machine must be compatible. The work area width is an integer ratio, indicating that the number of work strips per field is the ratio of the field width to the work area width. This can save unnecessary non-operational distances. When designing path sequence variables, Indicates a ridging machine The complete sequence of work routes, including in-field work routes and empty transfer routes between fields. Indicates the starting point, which is fixed at the warehouse location. To indicate the endpoint, it can be... It could also be other warehouses. Indicates the first in the path There are several nodes, which can be field entry points, field exit points, or warehouse points. When designing time variables, Indicates a ridging machine Start in the field The absolute time for the assignment, Indicates a ridging machine End in the field The absolute time for the assignment, Indicates a ridging machine Total working time. When designing soil compaction variables. Indicates all points in the field operation area The average number of times the object is run over by all the tires of the ridging machine.

[0056] Next, in stage three, the objective function is constructed: minimizing the total soil compaction degree. Compaction degree function Defined as the weighted sum of the number of compaction cycles across all field areas; the squared term is used to penalize areas with high compaction cycles, guiding the optimization algorithm to actively distribute paths and avoid repeated compaction. Minimize the total operation time: Total work time This includes the sum of the operating time, empty transfer time, and waiting time of all ridging machines. Overall objective function: ; For the comprehensive cost function, This represents the total soil compaction degree. Total operation time Weights based on decision-maker preferences. , A value close to 1 indicates a greater focus on reducing compaction; a value close to 0 indicates a greater focus on operational efficiency. It is a time constant. To improve compaction sensitivity, The set of coordinates for all fields. This is the average of the total number of times all points within the field operation area are rolled by the tires of all ridging machines; For ridging machine assembly, For ridging machine The time to start the assignment. For ridging machine The time to finish the task.

[0057] Phase four involves adding constraints, including task allocation integrity constraints. This indicates that each field must have exactly one ridging machine assigned to it; the field width and the working width are integer programming constraints: This represents the total number of ridges that a ridging machine can create on a given field, sufficient to cover the width of the field; minimum number of repeated compaction cycles constraint: Ridging machine operation path sequence and time logic constraints: This constraint describes that the operation sequence of the ridging machine must be sequential, and the start time of the later task must be later than the end time of the previous task plus the transfer time. For example, the ridging machine... In two consecutive fields and The time constraint for the transition between them is: ;in, For ridging machine Leave the field Time at the exit point For ridging machine Arrival at the field Entry point time, For fields The coordinates of the exit point, For fields The coordinates of the entry point, For ridging machine The empty-run transfer speed. Operation time window constraints: This constraint stipulates that all operations of each ridging machine must be completed within its designated time window, ensuring that ridging machines are not scheduled during unavailable periods (such as nighttime or maintenance periods); Operating cost constraint: .in, Indicates a ridging machine In the field Whether or not to perform the task is a decision variable; For the collection of fields; Indicates allocation to the ridging machine In the field The number of furrows on which the work needs to be completed; For ridging machine The working width; For fields The width; For the indicator function, if point Located in the ridging machine From node arrive If the value is 1 on the path segment, it is 0 otherwise. Nodes include field entrance point, field exit point and warehouse point. For ridging machine Reaching the node Time; For ridging machine Leave node Time; Represents a node and The distance the ridging machine travels between them; Indicates a ridging machine From node Move to The speed of travel; For ridging machine The starting point of the workable time window, For ridging machine Time of leaving the warehouse For ridging machine The time to return to the warehouse For ridging machine The end of the workable time window; For ridging machine Total operation time For ridging machine Unit time operation cost For the total budget.

[0058] Based on the above embodiments, this embodiment also provides a specific method for calling a preset algorithm to solve the problem. In the parameter initialization step, Indicates time step state, These represent the parameters of the policy network and the value network, respectively. Responsible for generating action probabilities, value network Responsible for assessing the status Good or bad. This represents the initial temperature of the simulated annealing. As a value greater than 0, a high temperature indicates a high probability of accepting inferior solutions, which is beneficial for global exploration; a low temperature indicates that it mainly accepts superior solutions, which is beneficial for local exploration. This represents the cooling coefficient for simulated annealing. After each training round, the temperature is updated to... . This represents an experience revisit cache, used to store experience tuples generated by the agent's interactions with the environment. ; This represents the optimal solution. In the environmental state acquisition step, This indicates that at the beginning of each training round or at each time step, the algorithm obtains the current state information from the simulation environment. The status of all ridging machines (location, working time, remaining budget, etc.). The status of all fields (whether operations have been completed, number of rows remaining, etc.). Indicates the distribution of soil compaction. This represents the time window constraint. In the forward propagation steps of the PPO policy network, Indicates the state Next, select any action. The probability distribution; Representing the value network in relation to state The value estimate represents the expected cumulative reward obtainable starting from this state. In the action sampling and simulated annealing optimization steps, the initial action sampling is as follows: This indicates that an initial action is sampled from the PPO probability distribution. Simulated annealing (SA) is used for local search, specifically neighborhood operations, through perturbation. Generate a new solution Then calculate the cost function. ; Indicates the action state The sum of time cost and soil compaction cost, This is the time cost weighting coefficient. To solidify the cost weighting coefficient, Action state Total homework time Action state Total soil compaction. The Metropolis criterion is... This indicates whether or not the decision has been made to accept the new interpretation. The principle, if Then always accept ,like Then, based on probability accept ;in, The new solution represents the action state. The sum of time cost and soil compaction cost, To simulate the current temperature parameters of the annealing algorithm, It is a natural constant.

[0059] In the steps of environmental interaction and experience storage, This represents interacting with the environment, receiving feedback, and storing experiences, where... The cost reward is calculated using the reward function. , As a cost incentive; This refers to the time-based reward weighting coefficient. For efficiency rewards, the value is a negative number representing the increase in total task time, meaning that shortening the total time will result in a positive reward; To solidify the reward weighting coefficient; As an ecological reward, the value is a negative number representing the increase in total soil compaction, meaning that reducing compaction will result in a positive reward; To constrain penalties, if the action Violation of any hard constraints (such as total budget overruns, conflicting workable time windows, etc.) It is a preset maximum negative reward, otherwise it is 0.

[0060] In the subsequent PPO network update step, the advantage function is estimated. ,in , As a discount factor, For GAE parameters; probability ratio ,in These are the network parameters before the update. The agent performs the action. Afterwards, you receive a reward for interacting with the environment. and new status experience tuple It has been updated and optimized.

[0061] After each round of training, lower the simulated annealing temperature and check the current strategy. Have you found a better one? A better solution is found, and if so, it is updated; the termination condition is met when the number of training rounds reaches the maximum value or the performance no longer improves within multiple consecutive rounds.

[0062] The ridging machine scheduling method for reducing soil mechanical compaction provided by this invention has been described in detail above. The various embodiments in the specification are described in a progressive manner, with each embodiment focusing on its differences from other embodiments. Similar or identical parts between embodiments can be referred to interchangeably. It should be noted that those skilled in the art can make various improvements and modifications to this invention without departing from its principles, and these improvements and modifications also fall within the protection scope of the claims of this invention.

Claims

1. A method for scheduling ridging machines to reduce soil mechanical compaction, characterized in that, include: Obtain information on field plots, warehouses, and ridging machines; A preset algorithm is invoked to solve the scheduling model based on the field information, warehouse information, and ridging machine information, with the goal of minimizing soil compaction and operation time, to obtain a ridging machine scheduling scheme.

2. The method for scheduling ridging machines to reduce soil mechanical compaction according to claim 1, characterized in that, The objective function of the scheduling model is: ； in, ; ； For the comprehensive cost function, This represents the total soil compaction degree. Total operation time Weights based on decision-maker preferences. It is a time constant. To improve compaction sensitivity, The set of coordinates for all fields. This is the average of the total number of times all points within the field operation area are rolled by the tires of all ridging machines; For ridging machine assembly, For ridging machine The time to start the assignment. For ridging machine The time to finish the task.

3. The method for scheduling ridging machines to reduce soil mechanical compaction according to claim 2, characterized in that, The constraints of the scheduling model include: ； ； ； ； ； ； in, Indicates a ridging machine In the field Whether or not to perform the task is a decision variable; For the collection of fields; Indicates allocation to the ridging machine In the field The number of furrows on which the work needs to be completed; For ridging machine The working width; For fields The width; For the indicator function, if point Located in the ridging machine From node arrive If the value is 1 on the path segment, it is 0 otherwise. The nodes include field entrance point, field exit point and warehouse point. For ridging machine Reaching the node Time; For ridging machine Leave node Time; Represents a node and The distance the ridging machine travels between them; Indicates a ridging machine From node Move to The speed of travel; For ridging machine The starting point of the workable time window, For ridging machine Time of leaving the warehouse For ridging machine The time to return to the warehouse For ridging machine The end of the workable time window; For ridging machine Total operation time For ridging machine Unit time operation cost For the total budget.

4. The method for scheduling ridging machines to reduce soil mechanical compaction according to claim 1, characterized in that, The preset algorithm is: PPO algorithm combined with simulated annealing algorithm; The invocation of the preset algorithm includes: conducting a global exploration using the PPO algorithm to make decisions on the allocation of tasks and path selection for the ridging machine; in the action selection stage of the PPO algorithm, using the simulated annealing algorithm to perturb the initial action to generate a new solution and calculate the cost function, and determining whether to accept the new solution based on the cost function.

5. The method for scheduling ridging machines to reduce soil mechanical compaction according to claim 4, characterized in that, The formula for calculating the cost function is as follows: ； Indicates the action state The sum of time cost and soil compaction cost, This is the time cost weighting coefficient. To solidify the cost weighting coefficient, Action state Total homework time Action state The total soil compaction degree.

6. The method for scheduling ridging machines to reduce soil mechanical compaction according to claim 5, characterized in that, The step of determining whether to accept a new solution based on the cost function includes: Calculate the difference in cost functions between the two actions: ; like Then always accept ,like Then, based on probability accept ; in, The new solution represents the action state. The sum of time cost and soil compaction cost, To simulate the current temperature parameters of the annealing algorithm, It is a natural constant.

7. The method for scheduling ridging machines to reduce soil mechanical compaction according to claim 6, characterized in that, In the interaction step with the environment in the PPO algorithm, the cost reward is calculated as follows: ； As a cost incentive; This refers to the time-based reward weighting coefficient. As an efficiency bonus, the value is a negative number representing the increase in total task time; To solidify the reward weighting coefficient; As an ecological reward, the value is a negative number representing the increase in total soil compaction. To constrain penalties, if the action Violation of any hard constraints It is a preset maximum negative reward, otherwise it is 0.

8. The method for scheduling ridging machines to reduce soil mechanical compaction according to claim 7, characterized in that, The hard constraints include: total budget overruns and workable time window conflicts.

9. The method for scheduling ridging machines to reduce soil mechanical compaction according to claim 3, characterized in that, Ridging machine In two consecutive fields and The time constraint for the transition between them is: ； in, For ridging machine Leave the field Time at the exit point For ridging machine Arrival at the field Entry point time, For fields The coordinates of the exit point, For fields The coordinates of the entry point, For ridging machine The empty transfer speed.

10. The method for scheduling ridging machines for reducing soil mechanical compaction according to any one of claims 1 to 9, characterized in that, The warehouse information includes: the number of warehouses, warehouse coordinates, and the maximum number of ridging machines that can be allocated to each warehouse; The field information includes: number of fields, field length, field width, minimum number of ridging strips per field, field entrance point, and field exit point; The information on the ridging machine includes: the number of ridging machines, the working width of the ridging machine, the field working speed, the empty transfer speed, the unit time operating cost, and the available working time window.

Images