Unmanned agricultural machine cluster path planning method based on multiple operation types

Abstract

The invention discloses an unmanned agricultural machine cluster path planning method based on multiple operation types, and relates to the technical field of agricultural automation, and the method comprises the steps: information acquisition and environment modeling: obtaining farmland, operation tasks and agricultural machine technical parameter information, and employing a grid method to construct a farmland environment model and operation models of different operation types; operation task planning: analyzing the operation tasks based on the operation model to determine the workload and the estimated time, determining the sequence of the operation types, and establishing a constraint rule, and carrying out the reasonable planning of the unmanned agricultural machinery cluster operation path through the comprehensive consideration of the sequence relation of different operation types. According to the method, unnecessary waiting time in the operation process of the agricultural machine is shortened, the phenomenon that the empty driving range is increased due to disordered operation sequences is avoided, meanwhile, in the path planning and optimizing link, multiple factors such as operation efficiency, cost and path length are comprehensively considered, the turning frequency and the running speed uniformity are optimized, and the path planning and optimizing efficiency is improved. And the energy consumption and abrasion of the agricultural machine are effectively reduced.

Description

Technical Field

[0001] This invention relates to the field of agricultural automation technology, specifically to a path planning method for unmanned agricultural machinery clusters based on multiple operation types. Background Technology

[0002] With the rapid advancement of technology, agricultural production is accelerating towards intelligence and automation. Under this trend, unmanned agricultural machinery cluster operations, with their advantages of high efficiency and precision, are gradually becoming a new direction for modern agricultural development. Actual agricultural production scenarios are complex and diverse, involving various types of operations such as tilling, sowing, fertilizing, weeding, and harvesting. Different types of operations have different requirements for the driving paths and operation sequences of agricultural machinery. Accurately planning the operation paths of unmanned agricultural machinery clusters is of great significance for improving agricultural production efficiency, reducing costs, and ensuring the quality of agricultural products.

[0003] Existing unmanned agricultural machinery path planning methods mostly have significant limitations and struggle to meet the practical needs of multi-tasking scenarios. On one hand, many existing methods are primarily designed for a single task type, considering only the machinery's path and operational efficiency within that task, while neglecting the interrelationships and sequential requirements between different task types. In actual agricultural production, there is often a strict order between different task types; for example, tilling must be performed first to loosen the soil before sowing. Reversing this order will severely impact crop growth and yield. Existing single-tasking path planning methods cannot coordinate the sequence of multiple task types, potentially leading to waiting times and repetitive tasks during operation, thus reducing efficiency. On the one hand, existing methods do not adequately consider resource constraints. Unmanned agricultural machinery consumes various resources during operation, such as pesticides, fertilizers, electricity, or fuel. When resources are insufficient, if they cannot be replenished in time, the operation will be interrupted. However, existing path planning methods often lack effective resource monitoring and early warning mechanisms, making it impossible to grasp the resource reserves of agricultural machinery in real time or adjust the operation path in a timely manner according to dynamic changes in resources. This results in resource waste and reduced operation efficiency. In addition, existing methods are also insufficient in dealing with complex and ever-changing farmland environments. They are difficult to adjust the path in real time according to factors such as farmland terrain and obstacle distribution, which may cause agricultural machinery to encounter obstacles or get stuck in unfavorable terrain during operation, affecting the smooth progress of the operation. Summary of the Invention

[0004] The purpose of this invention is to overcome the shortcomings of existing technologies and provide a path planning method for unmanned agricultural machinery clusters based on multiple operation types. It can obtain relevant information about farmland, operation tasks, and agricultural machinery through various means, and complete environmental modeling and operation model establishment; then, it performs operation task planning, clarifying the workload and operation sequence; subsequently, it carries out path planning and optimization, comprehensively considering multiple factors; during the process, it detects and resolves agricultural machinery conflicts in real time; finally, it dynamically adjusts the path based on resource monitoring and changes in the farmland environment.

[0005] To solve the above-mentioned technical problems, this invention provides the following technical solution: a path planning method for unmanned agricultural machinery clusters based on multiple operation types, the method comprising: Information Acquisition and Environmental Modeling: Acquire information on farmland, work tasks, and agricultural machinery technical parameters, and use the grid method to construct farmland environmental models and work models for different work types; Task planning: Based on the task model, analyze the task to determine the workload and estimated time, clarify the order of task types and establish constraint rules, and equip agricultural machinery with resource monitoring devices to monitor resource reserves in real time and set up early warning mechanisms; Path planning and optimization: Based on the defined work tasks, work sequence constraints and resource monitoring mechanisms, path planning is carried out, and the initial path is optimized by weight balancing to reduce unnecessary turns and ensure stable operation of agricultural machinery; Conflict detection and resolution: During the path planning process, conflicts in the agricultural machinery path are detected in real time. If a conflict occurs, the avoidance order is determined according to the operation priority and the avoidance agricultural machinery path is adjusted to bypass the conflict area. Dynamic route adjustment: Based on resource monitoring results and conflict handling, when a resource shortage warning is received or changes in the environment or agricultural machinery status are detected, the agricultural machinery route is dynamically adjusted, including the planned route to the supply point and the replanning or partial adjustment of the route.

[0006] Furthermore, in the information acquisition and environmental modeling steps, farmland information, including boundaries, terrain, and soil conditions, is acquired through sensors, satellite positioning systems, and farm management systems; operational task information, including specific requirements and priorities for different operational types, is acquired; and the technical parameters of each agricultural machine in the unmanned agricultural machinery cluster, including operating speed, operating width, and turning radius, are acquired. Based on the acquired information, a grid method is used to model the farmland environment, dividing the farmland into grid units of equal size. According to the farmland terrain and obstacle distribution, each grid unit is assigned attributes such as passable, impassable, and requiring specific operations. Operational models are established for different operational types. In the tillage operation model, the tillage depth requirements of agricultural machinery and soil hardness are considered to determine the operating energy consumption and operating time of agricultural machinery in each grid unit. In the sowing operation model, the sowing depth and row spacing are determined according to the seed variety and sowing requirements, and the correspondence between the sowing operation path and the grid unit is established.

[0007] Furthermore, in the task planning step, based on the established farmland environment model and task model, the task is analyzed. The workload and estimated time for each task are determined according to the task type, farmland area, and time requirements. Priority is also determined for different task types based on agronomic requirements. Simultaneously, each machine in the unmanned agricultural machinery cluster is equipped with a resource monitoring sensor to monitor its resource reserves in real time, including pesticide, fertilizer, and battery or fuel levels. The monitoring data is transmitted wirelessly to the cluster control center in real time. The cluster control center processes and analyzes the received resource monitoring data, sets resource warning thresholds, and issues a resource shortage warning signal when a machine's resource reserves fall below the threshold.

[0008] Furthermore, in the task planning step, the priority order among different task types is determined according to the agronomic requirements of agricultural production. The priority determination formula is as follows: ,in It is the first The priority of each task is used to determine the execution order of the tasks. It is the first The urgency coefficient for each task is determined based on agricultural time requirements, and its value ranges from [0, 1], with higher values ​​for more urgent tasks. It is the first The time requirement coefficient for each task is determined based on the time limit for completing the task, and its value ranges from [0, 1]. The stricter the time limit, the larger the value. It is the first The resource dependency coefficient of each task is determined based on the degree of dependence on resources for completing the task, and its value ranges from [0, 1]. The higher the degree of dependence, the larger the value. ε These are the weighting coefficients for urgency, time requirement, and resource dependence, respectively.

[0009] Furthermore, in the path planning and optimization step, path planning is carried out based on the determined work tasks, work sequence constraints, and resource monitoring mechanisms. The planning content includes the grid cell number, work type information, and work sequence information passed by each agricultural machine in different work stages. During the planning process, the path is evaluated by a path comprehensive evaluation index, taking into account work efficiency, work cost, path length, work sequence constraints, and resource constraints. By setting different weights to balance the influence of these factors, the initially planned path is optimized to reduce unnecessary turns during agricultural machine operations and enable the agricultural machine to travel at a stable speed.

[0010] Furthermore, in the path planning and optimization steps, the planning process utilizes a comprehensive path evaluation index. The path is evaluated, among which For the first The comprehensive evaluation index of the route, For the first The actual length of the path, This represents the maximum possible path length within the work area. For the first The actual operating cost of the route, This represents the maximum possible operating cost within the operating area. For the first The smoothness coefficient of a path represents the ease with which agricultural machinery can operate on that path. A higher value is taken when there are fewer turns and the travel speed is stable. For the first The degree of conformity of the operation sequence along the path. , , , These are the weighting coefficients for path length, operation cost, operation smoothness, and operation sequence conformity, respectively.

[0011] Furthermore, in the conflict detection and resolution step, during the path planning and optimization process, path conflicts between agricultural machines are detected in real time. By establishing a position prediction rule for agricultural machines, the position of agricultural machines in the future is predicted based on their speed and direction. The conflict risk is calculated using a conflict risk calculation formula. When a conflict is detected, a conflict resolution method based on priority and avoidance strategies is adopted. The avoidance order of conflicting agricultural machines is determined according to the priority of the work task, and the path of the avoiding agricultural machines is adjusted to bypass the conflict area. A feasible alternative path is found near the conflict area, and the alternative path meets the requirements of the work type and the technical parameter limitations of the agricultural machines.

[0012] Furthermore, in the conflict detection and resolution step, the conflict risk degree is calculated using the following formula: Calculate the conflict risk level, where Assess the risk of conflict between agricultural machinery. This refers to the real-time distance between agricultural machines. The safe distance threshold is determined based on the technical parameters of the agricultural machinery, such as its operating width and turning radius. The relative speed between agricultural machines This is the maximum speed of the agricultural machinery. , The weighting coefficients for distance and relative velocity are respectively (and ),like If the threshold is exceeded, a conflict is determined to have occurred.

[0013] Furthermore, in the dynamic path adjustment step, based on resource monitoring results and conflict handling, when a resource shortage warning signal is received, the agricultural machinery path is dynamically adjusted. Resource supply points are determined in the farmland and surrounding areas. The optimal path from the current location of the agricultural machinery to the nearest resource supply point, as well as the optimal path from the resource supply point to the subsequent work area, are calculated. These two paths are integrated with the original path to form a new work path. The path adjustment range is determined by a path adjustment range calculation formula. Based on real-time changes in the farmland environment, agricultural machinery status information, and dynamic resource monitoring results, the path is dynamically adjusted, and path planning or partial path adjustment is performed again.

[0014] Furthermore, in the path dynamic adjustment step, the path adjustment magnitude is calculated using the formula... Determine the path adjustment range, among which The adjustment range for the path. The degree of impact of changes in the farmland environment, To assess the impact of changes in resource reserves, the smaller the difference between the remaining resource level and the early warning threshold, the larger the value should be. This is the impact coefficient on path coherence after conflict resolution. , , These are the weighting coefficients for environmental change, resource change, and path coherence, respectively.

[0015] Compared with existing technologies, this unmanned agricultural machinery cluster path planning method based on multiple operation types has the following advantages: I. This invention comprehensively considers the sequential relationships of different operation types and rationally plans the operation path of unmanned agricultural machinery clusters. This reduces unnecessary waiting time for agricultural machinery during operation and avoids increased empty mileage caused by chaotic operation sequences. At the same time, in the path planning and optimization stage, it comprehensively considers factors such as operation efficiency, cost, and path length. By optimizing the number of turns and the uniformity of driving speed, it effectively reduces the energy consumption and wear of agricultural machinery, avoids repetitive operations and resource waste, significantly improves the operation efficiency of unmanned agricultural machinery clusters, increases resource utilization, and reduces agricultural production costs.

[0016] Second, this invention fully considers various factors such as farmland boundaries, topography, and soil conditions during the information acquisition and environmental modeling stages, establishing corresponding models for different operation types, making the planning method widely applicable. Moreover, by monitoring resource reserves in real time, the agricultural machinery path is dynamically adjusted in a timely manner when resources are insufficient, guiding the agricultural machinery to resource replenishment points and replanning subsequent operation paths. At the same time, the operation is carried out in strict accordance with the agronomic requirements, effectively avoiding operation interruptions, ensuring the smooth progress of the entire operation process, and improving the intelligence level and application scope of unmanned agricultural machinery clusters.

[0017] Other advantages, objectives and features of the invention will be set forth in part in the description which follows, and in part will be apparent to those skilled in the art from the following examination or study, or may be learned from the practice of the invention. Attached Figure Description

[0018] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the accompanying drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are merely some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without any creative effort.

[0019] Figure 1 A flowchart of a path planning method for unmanned agricultural machinery clusters based on multiple operation types; Figure 2 This is a flowchart illustrating the path planning and optimization process for a multi-tasking unmanned agricultural machinery cluster path planning method. Figure 3 This is a flowchart illustrating the conflict detection and resolution process for a path planning method for unmanned agricultural machinery clusters based on multiple operation types. Detailed Implementation

[0020] To further illustrate the technical means and effects of the present invention in achieving its intended purpose, the following detailed description of the specific implementation methods, structures, features, and effects of the present invention, in conjunction with the accompanying drawings and preferred embodiments, is provided below. Example

[0021] For multi-operation path planning in large farms in plains areas: like Figure 1As shown, information acquisition and environmental modeling were performed as follows: Using a satellite positioning system and a farm management system, information was acquired regarding the boundaries (rectangular, 1000 meters long and 667 meters wide), topography (flat terrain, maximum slope not exceeding 3 degrees, loam soil with moderate fertility), and soil conditions (pH 7.2, organic matter content 2.5%) of a 1000-mu contiguous farmland in a plain area. This identified three types of operations required: tillage, sowing, and fertilization. Tillage required a depth of 20 cm, and soil clods needed to be broken up to a diameter not exceeding 5 cm. Winter wheat was to be sown, and NPK compound fertilizer was to be applied. Information on 5 large tillage machines (operating speed 8-12 km / h, operating width 3 meters, minimum turning radius 5 meters) and 3 seeders (operating speed...) was also acquired. The technical parameters of the system are as follows: 1) a speed of 6-10 km / h, 2.4 m operating width, capable of simultaneously sowing 6 rows; 2) a fertilizer applicator (operating speed of 10-15 km / h, 4 m operating width). The farmland is divided into 10 m × 10 m grid units using a grid method, totaling 6670 grids. Based on the terrain and a few ditches (approximately 1 m wide and 0.8 m deep) obstacles, grid attributes are marked as passable, impassable, and corresponding operation types. Grids containing ditches are marked as impassable. Models are established for the three operation types. For example, the cultivated land model combines soil hardness to determine the energy consumption and operation time of each grid. The sowing model establishes the correspondence between the sowing path of each row and the grid based on the row spacing, ensuring that the spacing error between adjacent sowing rows within the grid does not exceed 2 cm.

[0022] Task planning: Based on environmental model analysis, the total workload for tillage is determined to be 1000 mu (approximately 67 hectares). It is estimated that each tillage machine can complete 80 mu (approximately 5 hectares) working 8 hours a day, requiring 3 days for 5 machines to operate simultaneously. The total workload for sowing is 1000 mu (approximately 67 hectares), with each seeder capable of completing 100 mu (approximately 6.7 hectares) per day, requiring 4 days for 3 machines to operate simultaneously. The total workload for fertilization is 1000 mu (approximately 67 hectares), with each fertilizer applicator capable of completing 200 mu (approximately 13 hectares) per day, requiring 3 days for 2 machines to operate simultaneously. The task priority calculation formula is used to determine the order of operations. Calculate the priority of each task. For the first Priority of each task. For the first The urgency coefficient of each task (arable land) ,sowing Fertilization ), For the first The time requirement coefficient for each task (all three are 0.7). For the first Resource dependence coefficient of each task (arable land) ,sowing Fertilization ), ε Using values ​​of 0.4, 0.3, and 0.3 respectively, the highest priority for cultivated land was calculated. At the same time, all agricultural machinery was equipped with high-precision sensors to monitor fuel levels, fertilizer levels, and seed levels in real time. Warnings were issued when fuel levels fell below 20% (approximately 50L) or fertilizer / seed levels fell below 30% of the amount required to complete the current area's work. The warning information was transmitted to the control center in real time via a 4G network.

[0023] Path planning and optimization: such as Figure 2 As shown, based on the task and sequence constraints, a path is planned for each agricultural machine, including the operation stage, grid number, and operation type. For example, tiller No. 1 is responsible for grids 1-1334 in the northwest area, and a straight east-west route is planned, with a distance of 3 meters between adjacent routes, which just covers its operation width. During the planning process, a comprehensive path evaluation index is used. Evaluate the path. For the first The comprehensive evaluation index of the route, For the first The actual length of the path, This represents the maximum possible path length within the work area (approximately 1500 meters). For the first The actual operating cost of the route, This represents the maximum possible operating cost within the work area (approximately 2000 yuan). For the first The smoothness coefficient of the operation of this path (this path) ), For the first Job sequence conformity of this path (this path) ), , , , The values ​​were set to 0.2, 0.3, 0.2, and 0.3 respectively, taking into account path length (prioritizing paths shorter than 1.2 times the diagonal of the area), cost (fuel consumption controlled within 0.3L / acre), smoothness (no more than 5 turns per kilometer), and sequence compliance (100% compliance with the work sequence). When optimizing the path, the turning angle of the tiller was controlled between 90° and 120° to avoid sharp turns and to keep the agricultural machinery traveling at a stable speed of 8-10 km / h, reducing extra fuel consumption caused by speed fluctuations.

[0024] Conflict detection and resolution: such as Figure 3As shown, the plan uses a real-time prediction method that updates the location of agricultural machinery once per second. It was found that tiller No. 3 (operating at grid 800) and seeder No. 1 (preparing to enter grid 750) may clash in 30 seconds at the northwest edge of the farmland. The clash area is grids 780-800. This clash is calculated using the clash risk degree formula. Calculate the degree of conflict risk. The degree of conflict risk between agricultural machinery (calculated in this case) (Exceeding the set threshold of 0.6), This represents the real-time distance between the agricultural machines (approximately 50 meters at this point). The safe distance threshold is set at 80 meters. The relative speed between the agricultural machines (approximately 2 meters per second). This refers to the maximum travel speed of the agricultural machinery (12 km / h, or 3.33 m / s). , The values ​​are set to 0.6 and 0.4 respectively. Based on the work priority (tillage takes priority over sowing), the control center sends an adjustment command to seeder No. 1, which temporarily detours to grid area No. 850. After tillage machine No. 3 passes through the conflict area, it will return to the original path. The detour distance is about 200 meters, which only increases the operation time by about 2 minutes.

[0025] Dynamic Path Adjustment: Towards the end of the operation, Fertilizer Applicator No. 2 (responsible for the northeast region) received a fertilizer shortage warning (only enough fertilizer remained to cover 50 mu, while 120 mu remained unfertilized). The control center immediately planned a path from its current location (grid number 450) to the nearest refueling station (located 300 meters east of the farmland). This path traversed 20 passable grids, was 500 meters long, and was expected to take 10 minutes to reach. Simultaneously, the optimal path from the refueling station to the remaining 120 mu (grid numbers 600-800) was planned to ensure continuous operation after refueling. This process was implemented through path adjustment calculations. Determine the path adjustment range. The adjustment range of the path (calculated this time) ), The degree of impact of changes in the farmland environment (at this point there is no significant change), ), The extent of the impact of changes in resource reserves (this resource change has a significant impact). ), The impact coefficient of path coherence after conflict resolution (no conflict in this case). ), , , The values ​​were 0.2, 0.6, and 0.2 respectively. When a short-term shower suddenly fell and lasted for 15 minutes, the soil sensor detected that the soil moisture in an area of ​​about 80 acres in the southwest corner of the farmland reached 35% (exceeding the suitable moisture level of 30% for sowing). The control center fine-tuned the path of Seeder No. 2 so that it would first work in other areas with suitable moisture. After 4 hours, when the moisture level in that area dropped to 28%, it would return to work to avoid uneven sowing depth caused by excessive moisture.

[0026] Example 2 For multi-operation path planning in small terraced fields in mountainous areas: Information Acquisition and Environmental Modeling: Aerial photography (using UAVs equipped with high-definition cameras and LiDAR) and ground surveying (using a total station) were conducted to acquire information on the boundaries (composed of 12 terraces of varying sizes, the largest being 8 mu and the smallest 2 mu) and topography (slope 15-25 degrees, with some areas having irregular obstacles due to exposed rocks, the largest rock diameter being 1.5 meters). This identified the areas requiring tillage (15 cm depth, with the inner side of the terraces 2 cm deeper than the outer side to facilitate water retention) and corn planting. Information was also acquired for two small tracked tillers (operating speed 3-5 km / h, operating width 1.2 m, minimum turning radius 2 m, track ground pressure 30 kPa, adaptable to steep slopes) and one small seeder (operating speed 2- The technical parameters (4 km / h, 0.8 m operating width, suitable for hill sowing) are used to model the operation. A variable-size grid (5 m × 5 m, densified to 3 m × 3 m at the edge of the terrace) is used, with a total of 1334 grids. The grid attributes are marked according to the terrace slope (15 degrees is marked as a light slope, 20-25 degrees as a moderate slope) and the distribution of stones (grids containing stones are marked as impassable). When establishing the operation model, the cultivated land model considers the impact of slope on the operation speed (5 km / h for light slopes, reduced to 3 km / h for moderate slopes) and energy consumption (energy consumption of moderate slopes is 20% higher than that of light slopes). The sowing model combines the direction of the terrace (along the contour lines) to determine the correspondence between the hill sowing position and the grid, and plans 8-10 sowing holes in each grid. Work task planning: After analysis, the total workload of tilling the land is determined to be 50 mu. Tiller No. 1 is responsible for 1-6 terraces (28 mu), and tiller No. 2 is responsible for 7-12 terraces (22 mu). It is estimated that each machine can complete 8 mu in 6 hours of work per day. The total workload of sowing is 50 mu, and it is estimated that 15 mu can be completed per day. The order is determined according to agronomic requirements: tilling → sowing. It is required that sowing must be carried out within 24 hours after tilling the same terrace. Sequence constraint rules are established. The agricultural machines are equipped with sensors to monitor fuel level (accuracy ±0.2L) and seed level (accuracy ±0.5kg). Warnings are set when the fuel level is below 25% (about 5L) and the seed level is below 25% (about 3kg). The monitoring data is transmitted through a wireless mesh network to ensure stable communication in mountainous areas with weak signals. Path planning and optimization: Based on the task-planned path, the impact of slope on the movement of agricultural machinery is given priority. All paths are planned along the contour lines of the terraces. For example, the path of the No. 1 tiller on the No. 3 terrace (slope of 20 degrees) is an "S" shape extending along the contour line, with a spacing of 1.2 meters between adjacent paths. When optimizing the turning path, the turning radius is controlled at 2-3 meters to adapt to the narrow areas of the terraces (only 4 meters at the narrowest point). A 0.5-meter buffer zone is set at the turning point to prevent the tracks of the agricultural machinery from running over the edge of the terrace and reduce the risk of rollover. At the same time, the operating speed is set at 4-5 km / h on light slopes and 3-4 km / h on moderate slopes. Conflict Detection and Resolution: On the morning of the second day of operation, it was detected that Tiller No. 1 (in Terrace No. 4, 60% complete) and Tiller No. 2 (in Terrace No. 5, 40% complete) might collide in one minute at the corner connecting the two terraces (grids 350-360). The corner was only 5 meters wide and could not accommodate two agricultural machines at the same time. Based on the operation progress, the control center instructed Tiller No. 2 to pause and wait at grid No. 340, and to enter idle mode. Tiller No. 1 (expected to pass through the corner in 2 minutes) would resume operation after it had completely passed. During the waiting period, only a small amount of fuel was consumed (about 0.1L). Dynamic Path Adjustment: On the afternoon of the 4th day of operation, at 3 PM, the No. 1 tiller (in Terrace No. 8) issued a low fuel warning (3L of fuel remaining, enough for only 1 hour of work, while 3 acres of the terrace remained uncultivated). The control center planned a path from its current position (grid No. 900) to the temporary refueling point on the mountainside (located 200 meters above Terrace No. 7). This path required traversing 15 gentle slope grids, with a length of 300 meters, and was expected to take 15 minutes to reach the point. Refueling time was approximately 5 minutes, after which the remaining work could be completed. On the morning of the 5th day, a small-scale rockfall occurred in the mountainous area. A drone inspection revealed 3 new rock obstacles on the edge of Terrace No. 5 (located in grids No. 500-520), with the largest rock having a diameter of 0.8 meters. The control center immediately made a partial adjustment to the seeder's path, causing it to detour around grid No. 530 to avoid the obstacle area. After the detour, the path length increased by 150 meters, and the working time increased by approximately 10 minutes, ensuring the seeder's safe passage.

[0027] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention in any way. Although the present invention has been disclosed above with reference to preferred embodiments, it is not intended to limit the present invention. Any person skilled in the art can make some modifications or alterations to the above-disclosed technical content to create equivalent embodiments without departing from the scope of the present invention. Any simple modifications, equivalent changes and alterations made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention shall still fall within the scope of the present invention.

Claims

1. A path planning method for unmanned agricultural machinery clusters based on multiple operation types, characterized in that, The method includes: Information Acquisition and Environmental Modeling: Acquire information on farmland, work tasks, and agricultural machinery technical parameters, and use the grid method to construct farmland environmental models and work models for different work types; Task planning: Based on the task model, analyze the task to determine the workload and estimated time, clarify the order of task types and establish constraint rules, and equip agricultural machinery with resource monitoring devices to monitor resource reserves in real time and set up early warning mechanisms; Path planning and optimization: Based on the defined work tasks, work sequence constraints and resource monitoring mechanisms, path planning is carried out, and the initial path is optimized by weight balancing to reduce unnecessary turns and ensure stable operation of agricultural machinery; Conflict detection and resolution: During the path planning process, conflicts in the agricultural machinery path are detected in real time. If a conflict occurs, the avoidance order is determined according to the operation priority and the avoidance agricultural machinery path is adjusted to bypass the conflict area. Dynamic route adjustment: Based on resource monitoring results and conflict handling, when a resource shortage warning is received or changes in the environment or agricultural machinery status are detected, the agricultural machinery route is dynamically adjusted, including the planned route to the supply point and the replanning or partial adjustment of the route.

2. The unmanned agricultural machinery cluster path planning method based on multiple operation types according to claim 1, characterized in that, In the information acquisition and environmental modeling steps, farmland information, including boundaries, terrain, and soil conditions, is acquired through sensors, satellite positioning systems, and farm management systems; operational task information, including specific requirements and priorities for different operational types, is acquired; and the technical parameters of each agricultural machine in the unmanned agricultural machinery cluster, including operating speed, operating width, and turning radius, are acquired. Based on the acquired information, a grid method is used to model the farmland environment, dividing the farmland into grid cells of equal size. According to the farmland terrain and obstacle distribution, each grid cell is assigned attributes such as passable, impassable, and requiring specific operations. Operational models are established for different operational types. In the tillage operation model, the tillage depth requirements of agricultural machinery and soil hardness are considered to determine the operating energy consumption and operating time of agricultural machinery in each grid cell. In the sowing operation model, the sowing depth and row spacing are determined according to the seed variety and sowing requirements, and the correspondence between the sowing operation path and the grid cell is established.

3. The unmanned agricultural machinery cluster path planning method based on multiple operation types according to claim 1, characterized in that, In the task planning step, based on the established farmland environment model and operation model, the task is analyzed. According to the task type, farmland area, and time requirements, the workload and estimated operation time of each task are determined. According to the agronomic requirements of agricultural production, the priority order between different task types is determined. At the same time, each machine in the unmanned agricultural machinery cluster is equipped with a resource monitoring sensor to monitor the resource reserves of the machine in real time, including pesticide reserves, fertilizer reserves, and power or fuel levels. The monitoring data is transmitted to the cluster control center in real time via wireless communication. The cluster control center processes and analyzes the received resource monitoring data, sets a resource warning threshold, and issues a resource shortage warning signal when the resource reserves of a machine fall below the warning threshold.

4. The unmanned agricultural machinery cluster path planning method based on multiple operation types according to claim 3, characterized in that, In the task planning step, the priority order among different task types is determined according to the agronomic requirements of agricultural production. The priority determination formula is as follows: ,in It is the first The priority of each task is No. The urgency coefficient of each task. It is the first The time requirement coefficient for each task. It is the first Resource dependency coefficient of each task ε These are the weighting coefficients for urgency, time requirement, and resource dependence, respectively.

5. The unmanned agricultural machinery cluster path planning method based on multiple operation types according to claim 1, characterized in that, In the path planning and optimization step, path planning is carried out based on the determined work tasks, work sequence constraints, and resource monitoring mechanisms. The planning content includes the grid cell number, work type information, and work sequence information that each agricultural machine passes through at different work stages. During the planning process, the path is evaluated by a path comprehensive evaluation index, taking into account work efficiency, work cost, path length, work sequence constraints, and resource constraints. By setting different weights to balance the influence of these factors, the initially planned path is optimized to reduce unnecessary turns during agricultural machine operations and enable the agricultural machine to travel at a stable speed.

6. The unmanned agricultural machinery cluster path planning method based on multiple operation types according to claim 5, characterized in that, In the path planning and optimization steps, the planning process uses a path comprehensive evaluation index. The path is evaluated, among which For the first The comprehensive evaluation index of the route, For the first The actual length of the path, This represents the maximum possible path length within the work area. For the first The actual operating cost of the route, This represents the maximum possible operating cost within the operating area. For the first The smoothness coefficient of the operation path, For the first The degree of conformity of the operation sequence along the path. , , , These are the weighting coefficients for path length, operation cost, operation smoothness, and operation sequence conformity, respectively.

7. The unmanned agricultural machinery cluster path planning method based on multiple operation types according to claim 1, characterized in that, In the conflict detection and resolution steps, during the path planning and optimization process, path conflicts between agricultural machinery are detected in real time. By establishing a position prediction rule for agricultural machinery, the position of agricultural machinery in the future is predicted based on the speed and direction of the agricultural machinery. The conflict risk is calculated using a conflict risk calculation formula. When a conflict is detected, a conflict resolution method based on priority and avoidance strategy is adopted. The avoidance order of conflicting agricultural machinery is determined according to the priority of the operation task, and the path of the avoiding agricultural machinery is adjusted to bypass the conflict area. A feasible alternative path is found near the conflict area. The alternative path meets the requirements of the operation type and the technical parameter restrictions of the agricultural machinery.

8. The unmanned agricultural machinery cluster path planning method based on multiple operation types according to claim 7, characterized in that, In the conflict detection and resolution steps, the conflict risk degree is calculated using the following formula: Calculate the conflict risk level, where Assess the risk of conflict between agricultural machinery. This refers to the real-time distance between agricultural machines. For safe distance threshold, The relative speed between agricultural machines This is the maximum speed of the agricultural machinery. , The weighting coefficients for distance and relative velocity are respectively (and ),like If the threshold is exceeded, a conflict is determined to have occurred.

9. The unmanned agricultural machinery cluster path planning method based on multiple operation types according to claim 1, characterized in that, In the dynamic path adjustment step, based on resource monitoring results and conflict handling, when a resource shortage warning signal is received, the agricultural machinery path is dynamically adjusted. Resource supply points are determined in the farmland and surrounding areas. The optimal path from the current location of the agricultural machinery to the nearest resource supply point and the optimal path from the resource supply point to the subsequent operation area are calculated. These two paths are integrated with the original path to form a new operation path. The path adjustment range is determined by the path adjustment range calculation formula. Based on real-time changes in the farmland environment, agricultural machinery status information, and dynamic resource monitoring results, the path is dynamically adjusted, and the path is re-planned or partially adjusted.

10. The unmanned agricultural machinery cluster path planning method based on multiple operation types according to claim 9, characterized in that, In the dynamic path adjustment step, the path adjustment range is determined by a path adjustment range calculation formula. Determine the path adjustment range, among which The adjustment range for the path. The degree of impact of changes in the farmland environment, To assess the impact of changes in resource reserves, the smaller the difference between the remaining resource level and the early warning threshold, the larger the value should be. This is the impact coefficient on path coherence after conflict resolution. , , These are the weighting coefficients for environmental change, resource change, and path coherence, respectively.

Images