Self-propelled sprayer scheduling method based on dual-objective optimization

Abstract

The invention relates to the technical field of agricultural machinery scheduling, discloses a self-propelled sprayer scheduling method based on dual-objective optimization, and aims to solve the problem of soil compaction caused by the fact that a sprayer repeatedly enters and exits a field during sprayer scheduling in the prior art. According to the application, the optimal sprayer scheduling scheme is sought by taking the minimum waiting time and the soil compaction degree deviation as targets, and through the pre-constructed scheduling model, according to the coordinates of the field needing to be operated, the coordinates of the warehouse where the schedulable sprayer is located and the operation information of the sprayer, the optimal sprayer scheduling scheme is obtained. A preset algorithm is adopted as a strategy to solve the scheduling model to obtain a sprayer scheduling scheme, reasonable scheduling of a sprayer cluster is achieved, and the influence on soil compaction is reduced.

Description

Technical Field

[0001] This invention relates to the field of agricultural machinery scheduling technology, and in particular to a scheduling method for self-propelled sprayers based on bi-objective optimization. Background Technology

[0002] The sustainable development of modern agriculture highly depends on the protection of farmland ecological environment and the maintenance of productivity. Among the many stages of crop production, soil mechanical compaction has become an increasingly serious global problem. Large agricultural machinery, including the widely used self-propelled high-clearance sprayers, transfers heavy loads to the soil through tires or tracks during field operations, causing soil particles to be compressed, pore structure to be destroyed, bulk density to increase, and aeration and water conductivity to decrease. This compaction effect not only directly damages soil health, inhibits root growth, and reduces water and nutrient use efficiency, but may also exacerbate surface runoff and soil erosion, forming a vicious cycle that ultimately threatens crop yield potential and long-term farmland productivity.

[0003] Modern large self-propelled sprayers are widely used in large-scale farms due to their high operating efficiency, large-capacity spray tanks, and excellent maneuverability. However, their significant weight and specific operating patterns (such as full coverage and frequent entry and exit from the field) make them a major source of soil compaction. The damage caused by compaction is particularly pronounced when operating during periods of high soil moisture and lower load-bearing capacity.

[0004] Therefore, how to reduce soil compaction while enabling sprayers to perform their operations 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 address the problem of soil compaction caused by repeated entry and exit of sprayers in the field when scheduling sprayers in the current technology. Therefore, this invention provides a self-propelled sprayer scheduling method based on dual-objective optimization to achieve reasonable scheduling of sprayer clusters and reduce the impact on soil compaction.

[0006] To address the aforementioned technical problems, this invention provides a self-propelled sprayer scheduling method based on dual-objective optimization, comprising:

[0007] Obtain the coordinates of fields and warehouses, as well as the operating information of sprayers;

[0008] The preset algorithm is invoked to minimize the waiting time and soil compaction deviation. Based on coordinate information and operation information, the scheduling model is solved to obtain the sprayer scheduling scheme.

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

[0010] ;

[0011] ;

[0012] ;

[0013] ;

[0014] in, Indicates the first The waiting time before the sprayer arrives at the field to carry out operations. For the first The deviation between the soil compaction degree after spraying and the target soil compaction degree. This is a weighting factor used to balance the waiting time and the deviation in soil compaction. For the first Soil compaction degree after spraying with a sprayer. For the target compaction degree of the soil, For the first The start time of the operation by the sprayer. For the first The departure time of the sprayer from the starting point. For the first The travel time of the sprayer from the starting point to the field. For the collection of sprayers, The shortest distance between the warehouse and the field. The operating speed of the sprayer;

[0015] The constraints include:

[0016] ;

[0017] ;

[0018] ;

[0019] ;

[0020] ;

[0021] ;

[0022] in, This represents the minimum soil compaction value for the sprayer. This represents the maximum soil compaction value achieved by the sprayer. For the first The operating path of the sprayer. For the first The operating path of the sprayer. For the first A sprayer from the field to the field The movement decision quantity, if the first A sprayer from the field Move to field ,but =1, otherwise =0, For the first Taiwan sprayer Time period in the field The intensity of pesticide application operations, For fields Minimum required application intensity For the first The maximum spraying intensity of the sprayer. For the collection of fields, It is a set of time periods.

[0023] Preferred methods for calculating the distance between the warehouse and the field include:

[0024] Get the road type indicator for warehouses and fields;

[0025] If the road type indicator is 0, then the road type is confirmed to be paved road. ;

[0026] If the road type indicator is 1, then the road type is confirmed to be an unpaved road. ;

[0027] in, ; ;

[0028] For fields coordinate, The coordinates of the warehouse are... This is the correction factor for dirt roads.

[0029] Preferably, the preset algorithm is a reinforcement learning algorithm combined with an ant colony algorithm. The process of calling the preset algorithm to solve the scheduling model to obtain the sprayer scheduling scheme includes:

[0030] The sprayer scheduling task is assigned using a reinforcement learning algorithm;

[0031] Based on the allocation results, ant paths are constructed using the ant colony algorithm, and the fitness of each path is calculated.

[0032] Pheromones are updated based on fitness to obtain the optimal sprayer scheduling scheme.

[0033] Preferably, when assigning sprayer scheduling tasks using a reinforcement learning algorithm, the formula for calculating the immediate reward factor is:

[0034] ;

[0035] in, ;

[0036] ; ;

[0037] in, As an instant reward factor, As the target weight for soil compaction, As the cumulative time target weight, To increase soil compaction volume, For the first Maximum soil compaction after spraying operation by a sprayer. To increase the waiting time, As the normalized base for the waiting time, For the first A sprayer in the field The increased time required for soil compaction operations in sensitive areas For homework time, For fields Soil volumetric water content, This represents the maximum soil volumetric water content of the field. Soil moisture factor, This refers to the application intensity of the sprayer in different fields.

[0038] Preferably, the first step is to extract the first step using a reinforcement learning algorithm. The state characteristics of the sprayer for task allocation in the feature space are as follows:

[0039] ;

[0040] For the first The status characteristics of the sprayer.

[0041] Preferably, the formula for calculating path fitness is:

[0042] ; This represents path fitness.

[0043] Preferably, the pheromone update formula is:

[0044] ;

[0045] ;

[0046] For fields to the field Ant pheromones, For the updated field to the field Ant pheromones For ant colony pheromone evaporation rate, To assign sprayer scheduling tasks using reinforcement learning algorithms Value table.

[0047] Preferably, when constructing ant paths, the formula for calculating the transition probability is:

[0048] ; ;

[0049] in, To reach the field Time, For heuristic factor coefficients, Let be the transition probability, representing the th Taiwan sprayer based on pheromones from the field Directly transferred to the field With from the field to the field The probability of the sum ratio of pheromones between two adjacent fields during the operation.

[0050] Preferably, after the pheromone update step, the method further includes: enhancing the pheromones through an elite strategy. ;

[0051] in, Enhancement coefficient for elites, , For enhanced ant pheromones.

[0052] This invention provides a self-propelled sprayer scheduling method based on dual-objective optimization. Compared with the current technology, which ignores the problem of soil compaction caused by repeated entry and exit of sprayers in the field when scheduling sprayers, this application seeks the optimal sprayer scheduling scheme by minimizing waiting time and soil compaction deviation as objectives. Through a pre-constructed scheduling model, based on the coordinates of the field to be operated, the coordinates of the warehouse where the schedulable sprayers are located, and the operating information of the sprayers, a preset algorithm is used as a strategy to solve the scheduling model to obtain the sprayer scheduling scheme, thereby realizing the reasonable scheduling of the sprayer cluster and reducing the impact on soil compaction. Attached Figure Description

[0053] 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.

[0054] Figure 1 The flowchart illustrates a self-propelled sprayer scheduling method based on dual-objective optimization, as provided in an embodiment of the present invention. Detailed Implementation

[0055] 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.

[0056] The core of this invention is to provide a self-propelled sprayer scheduling method based on dual-objective optimization, which can achieve reasonable scheduling of sprayer clusters and reduce the impact on soil compaction.

[0057] 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.

[0058] Figure 1 A flowchart of a self-propelled sprayer scheduling method based on dual-objective optimization is provided for an embodiment of the present invention, as shown below. Figure 1 As shown, the method includes:

[0059] S10: Obtain the coordinates of the fields and warehouses, as well as the operating information of the sprayers;

[0060] S11: Call the preset algorithm to minimize waiting time and soil compaction deviation, and solve the scheduling model based on coordinate information and operation information to obtain the sprayer scheduling scheme.

[0061] The self-propelled sprayer scheduling method based on dual-objective optimization provided in this application is mainly used to schedule self-propelled sprayers in smart agriculture, reducing operation time and minimizing soil compaction damage while ensuring operation. In specific implementations, the execution entity of this method can be a self-propelled sprayer scheduling device based on dual-objective optimization. This device may 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 self-propelled sprayer scheduling method based on dual-objective optimization provided in this application. This device can cooperate with the devices on the sprayers to allocate tasks to each sprayer. In some embodiments, the self-propelled sprayer scheduling device based on dual-objective optimization may also include human-computer interaction devices such as displays and touchscreens. In specific implementations, the self-propelled sprayer scheduling device based on dual-objective optimization provided in this embodiment may include, but is not limited to, smartphones, tablets, laptops, or desktop computers.

[0062] 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.

[0063] In this application, the scheduling and allocation of sprayers is based on a scheduling model. By collecting data related to the model, a preset algorithm is invoked to allocate sprayers according to a certain strategy. The purpose of this application is to reduce the operating time of sprayers and soil compaction. The main factor affecting the operating time is the waiting time before the sprayer travels to the field to perform the operation. The degree of impact caused by soil compaction can be described by the soil compaction deviation. Therefore, the scheduling model can be constructed with the goal of minimizing the waiting time and soil compaction deviation. The parameters used in the construction process are related to the field location, the sprayer location, and the sprayer's operating parameters such as travel speed. In specific implementation, the field location and the sprayer location can be represented by coordinates. Since sprayers are often uniformly placed in a warehouse, the location of the sprayers can be unified by the warehouse coordinates. Therefore, in practical use, this application can obtain the coordinate information of the field and warehouse, as well as the operating information of the sprayer, through step S10, and then perform subsequent model solving. The sprayer's operational information is primarily used to calculate its waiting time. This value is related to the sprayer's operating speed and the start time of its field operations. Therefore, this operational information includes the sprayer's travel speed, operating time, and operating intensity. Subsequently, by invoking a preset algorithm, with the goal of minimizing waiting time and soil compaction deviation, the scheduling model is solved based on coordinate and operational information to obtain the sprayer scheduling scheme.

[0064] This embodiment also provides a specific method for constructing a scheduling model, thereby obtaining a sprayer scheduling scheme by solving the model using a preset algorithm and a certain strategy. The objective function of the scheduling model is:

[0065] ;

[0066] ;

[0067] ;

[0068] ;

[0069] in, Indicates the first The waiting time before the sprayer arrives at the field to carry out operations. For the first The deviation between the soil compaction degree after spraying and the target soil compaction degree. This is a weighting factor used to balance the waiting time and the deviation in soil compaction. For the first Soil compaction degree after spraying with a sprayer. For the target compaction degree of the soil, For the first The start time of the operation by the sprayer. For the first The departure time of the sprayer from the starting point. For the first The travel time of the sprayer from the starting point to the field. For the collection of sprayers, The shortest distance between the warehouse and the field. The operating speed of the sprayer;

[0070] The constraints include:

[0071] ;

[0072] ;

[0073] ;

[0074] ;

[0075] ;

[0076] ;

[0077] in, This represents the minimum soil compaction value for the sprayer. This represents the maximum soil compaction value achieved by the sprayer. For the first The operating path of the sprayer. For the first The operating path of the sprayer. For the first A sprayer from the field to the field The movement decision quantity, if the first A sprayer from the field Move to field ,but =1, otherwise =0, For the first Taiwan sprayer Time period in the field The intensity of pesticide application operations, For fields Minimum required application intensity For the first The maximum spraying intensity of the sprayer. For the collection of fields, It is a set of time periods.

[0078] The self-propelled sprayer scheduling method based on dual-objective optimization provided in this application addresses the issue of soil compaction caused by repeated entry and exit of sprayers in the field, which is ignored in current technologies. This application seeks the optimal sprayer scheduling scheme by minimizing waiting time and soil compaction deviation. Through a pre-constructed scheduling model, the scheduling scheme is obtained by solving the scheduling model with a preset algorithm based on the coordinates of the field to be operated, the coordinates of the warehouse where the schedulable sprayers are located, and the operating information of the sprayers. This achieves reasonable scheduling of the sprayer cluster and reduces the impact on soil compaction.

[0079] It is understandable that the distance between the warehouse and the field can be directly obtained from the warehouse coordinates and the field coordinates. Based on the above embodiments, this application considers that in actual application scenarios, the locations of the warehouse and the field differ, and the difficulty of traveling between them varies, which can lead to errors in subsequent calculations of travel time. Therefore, to avoid these errors, this embodiment performs a unified distance calculation based on different road conditions. Specifically, it identifies a road type indicator used to indicate the road between the warehouse and the field. This indicator has two states: 0 and 1. If the road type indicator is 0, then the road type is confirmed to be a paved road. If the road type indicator is 1, then the road type is confirmed to be an unpaved road. ;in, ; ; For fields coordinate, The coordinates of the warehouse are... This is the correction factor for dirt roads.

[0080] Building upon the technology of the above embodiments, this embodiment provides a specific preset algorithm that uses a reinforcement learning algorithm combined with an ant colony algorithm (RL-ACO). The algorithm utilizes the value function of the reinforcement learning algorithm to correct the pheromone update rules of the ant colony algorithm, allowing ants to prioritize paths with "high value." Specifically, the process of calling the preset algorithm to solve the scheduling model and obtain the sprayer scheduling scheme includes: allocating sprayer scheduling tasks using a reinforcement learning algorithm; based on the allocation results, constructing ant paths using the ant colony algorithm and calculating the fitness of each path; and updating the pheromone based on the fitness to obtain the optimal sprayer scheduling scheme.

[0081] This embodiment provides a specific reinforcement learning algorithm, using the Q-learning algorithm, through... The value table modifies the pheromone update rule of the ant colony algorithm. Specifically, when allocating sprayer scheduling tasks using the reinforcement learning algorithm, the formula for calculating the immediate reward factor is:

[0082] ;

[0083] in, ;

[0084] ; ;

[0085] in, As an instant reward factor, As the target weight for soil compaction, As the cumulative time target weight, To increase soil compaction volume, For the first Maximum soil compaction after spraying operation by a sprayer. To increase the waiting time, As the normalized base for the waiting time, For the first A sprayer in the field The increased time required for soil compaction operations in sensitive areas For homework time, For fields Soil volumetric water content, This represents the maximum soil volumetric water content of the field. Soil moisture factor, This refers to the application intensity of the sprayer in different fields.

[0086] Extracting the first through reinforcement learning algorithm The state characteristics of the sprayer for task allocation in the feature space are as follows:

[0087] ;

[0088] For the first The status characteristics of the sprayer.

[0089] When selecting ε-greedy actions, ,in For exploration rate, , To make actionable collections. This represents the current state of the sprayer in the field being worked on. This refers to the operating status of the sprayer. for The choice of a similar, better action state indicates the probability of exploration when choosing. At that time, the action Choose one that reflects the normal range of motion. Value, when the probability is When, choose with Similar action states of value.

[0090] Next, determine the actions to be taken when the sprayer is assigned to a field and update the status: , The fitness parameter for the sprayer scheduling task allocation state, when the ε-greedy action selects a suitable action state. Then, the action status is updated to adapt to the new sprayer assignment tasks.

[0091] The updated reinforcement learning can be obtained by calculating the reward factor. Value table, .

[0092] Then, the ant colony algorithm can be used for path optimization. First, the pheromone is initialized. , For fields to the field Ant pheromones, The initial pheromone concentration is used, followed by ant path construction. , No. An ant represents a sprayer. This formula represents a sprayer based on ant information. Ant paths constructed under job path constraints.

[0093] When constructing ant paths, the transition probability is calculated as follows:

[0094] ; ;

[0095] in, To reach the field Time, For heuristic factor coefficients, Let be the transition probability, representing the th Taiwan sprayer based on pheromones from the field Directly transferred to the field With from the field to the field The probability of the sum ratio of pheromones between two adjacent fields during the operation.

[0096] After calculating the transition probability, the next node is selected, and the path time is updated. After path construction is complete, the path fitness is calculated using the following formula: ; This represents path fitness.

[0097] The pheromone is updated based on path fitness, specifically as follows:

[0098] ;

[0099] ;

[0100] For fields to the field Ant pheromones, For the updated field to the field Ant pheromones For ant colony pheromone evaporation rate, To assign sprayer scheduling tasks using reinforcement learning algorithms Value table.

[0101] Subsequently, in order to obtain the optimal solution, after obtaining the updated pheromones, an elite strategy was used to enhance the pheromones. ;in, Enhancement coefficient for elites, , , For enhanced ant pheromones.

[0102] Finally, when the iteration reaches its maximum value, the optimal solution is output: , This represents the set of optimal paths for reducing compaction and waiting time. This indicates the optimal allocation scheme for sprayer scheduling in the field.

[0103] The foregoing has provided a detailed description of the self-propelled sprayer scheduling method based on dual-objective optimization provided by this invention. 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 self-propelled sprayer scheduling method based on dual-objective optimization, characterized in that, include: Obtain the coordinates of fields and warehouses, as well as the operating information of sprayers; The preset algorithm is invoked to minimize the waiting time and soil compaction deviation. Based on coordinate information and operation information, the scheduling model is solved to obtain the sprayer scheduling scheme.

2. The self-propelled sprayer scheduling method based on dual-objective optimization according to claim 1, characterized in that, The objective function of the scheduling model is: ； ； ； ； in, Indicates the first The waiting time before the sprayer arrives at the field to carry out operations. For the first The deviation between the soil compaction degree after spraying and the target soil compaction degree. This is a weighting factor used to balance the waiting time and the deviation in soil compaction. For the first Soil compaction degree after spraying operation by a sprayer. For the target compaction degree of the soil, For the first The start time of the operation by the sprayer. For the first The departure time of the sprayer from the starting point. For the first The travel time of the sprayer from the starting point to the field. For the collection of sprayers, The shortest distance between the warehouse and the field. The operating speed of the sprayer; The constraints include: ； ； ； ； ； ； in, This represents the minimum soil compaction value achieved by the sprayer. This represents the maximum soil compaction value achieved by the sprayer. For the first The operating path of the sprayer. For the first The operating path of the sprayer. For the first A sprayer from the field to the field The movement decision quantity, if the first A sprayer from the field Move to field ,but =1, otherwise =0, For the first Taiwan sprayer Time period in the field The intensity of pesticide application operations, For fields Minimum required application intensity For the first The maximum spraying intensity of the sprayer. For the collection of fields, It is a set of time periods.

3. The self-propelled sprayer scheduling method based on dual-objective optimization according to claim 2, characterized in that, The methods for calculating the distance between the warehouse and the field include: Get the road type indicator for warehouses and fields; If the road type indicator is 0, then the road type is confirmed to be paved road. ; If the road type indicator is 1, then the road type is confirmed to be an unpaved road. ; in, ; ; For fields coordinate, The coordinates of the warehouse are... This is the correction factor for dirt roads.

4. The self-propelled sprayer scheduling method based on dual-objective optimization according to claim 3, characterized in that, The preset algorithm is a reinforcement learning algorithm combined with an ant colony algorithm. The preset algorithm is used to solve the scheduling model to obtain the sprayer scheduling scheme, including: The sprayer scheduling task is assigned using a reinforcement learning algorithm; Based on the allocation results, ant paths are constructed using the ant colony algorithm, and the fitness of each path is calculated. Pheromones are updated based on fitness to obtain the optimal sprayer scheduling scheme.

5. The self-propelled sprayer scheduling method based on dual-objective optimization according to claim 4, characterized in that, When assigning sprayer scheduling tasks using reinforcement learning algorithms, the formula for calculating the immediate reward factor is: ； in, ; ； ； in, As an instant reward factor, As the target weight for soil compaction, As the cumulative time target weight, To increase soil compaction volume, For the first Maximum soil compaction after spraying operation by a sprayer. To increase the waiting time, As the normalized base for the waiting time, For the first A sprayer in the field The increased time required for soil compaction operations in sensitive areas For homework time, For fields Soil volumetric water content, This represents the maximum soil volumetric water content of the field. Soil moisture factor, This refers to the application intensity of the sprayer in different fields.

6. The self-propelled sprayer scheduling method based on dual-objective optimization according to claim 5, characterized in that, Extracting the first through reinforcement learning algorithm The state characteristics of the sprayer for task allocation in the feature space are as follows: ； For the first The status characteristics of the sprayer.

7. The self-propelled sprayer scheduling method based on dual-objective optimization according to claim 6, characterized in that, The formula for calculating path fitness is: ; This represents path fitness.

8. The self-propelled sprayer scheduling method based on dual-objective optimization according to claim 7, characterized in that, The pheromone update formula is: ； ； For fields to the field Ant pheromones, For the updated field to the field Ant pheromones For ant colony pheromone evaporation rate, To assign sprayer scheduling tasks using reinforcement learning algorithms Value table.

9. The self-propelled sprayer scheduling method based on dual-objective optimization according to claim 8, characterized in that, When constructing ant paths, the transition probability is calculated as follows: ； ； in, To reach the field Time, For heuristic factor coefficients, Let be the transition probability, representing the th Taiwan sprayer based on pheromones from the field Directly transferred to the field With from the field to the field The probability of the ratio of the sum of pheromones of two adjacent fields during the operation.

10. The self-propelled sprayer scheduling method based on dual-objective optimization according to claim 8, characterized in that, Following the pheromone update step, the process also includes: enhancing pheromones through elite strategies. ; in, Enhancement coefficient for elites, , For enhanced ant pheromones.

Images