Unmanned agricultural machine real-time obstacle avoidance trajectory planning method, equipment and medium

Abstract

The invention relates to the technical field of unmanned agricultural machine path planning, relates to the technical field of aircraft falling speed constraint guidance, provides an unmanned agricultural machine real-time obstacle avoidance trajectory planning method and device and a medium, and aims to realize unmanned agricultural machine real-time obstacle avoidance trajectory planning based on a multi-strategy fusion improved colistification algorithm. The method is used for obtaining the smooth, stable, safe and collision-free optimal track meeting the kinematics and dynamics constraints of the agricultural machine in real time in a complex dynamic environment. The method overcomes the defects of a traditional planning algorithm in the aspects of calculation efficiency, adaptability, robustness and the like, and improves the trajectory planning effect.

Description

Technical Field

[0001] The present invention relates to the technical field of unmanned agricultural machinery path planning, and in particular to a method, device and medium for real-time obstacle avoidance trajectory planning for unmanned agricultural machinery. Background Art

[0002] Mechanized farming roads serve as crucial infrastructure for autonomous agricultural machinery to travel between hangars and farmland, facilitating full-scale mechanized operations and large-scale production. Unlike structured roads, these roads are semi-structured and lack markings. Their complex and ever-changing environments often feature a wide variety of dynamic and static obstacles and traffic participants. Therefore, to meet the requirements of autonomous agricultural machinery operations, agricultural input transportation, and other agricultural production activities, implementing real-time local path planning and autonomous obstacle avoidance for autonomous agricultural machinery within these roads is of great practical significance.

[0003] Trajectory planning is a typical three-dimensional constraint planning problem that considers both time and space. It outputs a series of trajectory points containing path and velocity information, taking into account environmental constraints, vehicle kinematics, and dynamics. Compared to path planning, trajectory planning with the added time constraint is computationally more complex. A common approach is to decouple path and velocity, first planning a series of optimal path points and then assigning a desired velocity value based on the optimal path. This method reduces the three-dimensional solution space to two dimensions, significantly reducing solution complexity. Common local path planning methods for agricultural machinery include artificial potential field methods, parametric curve methods, and numerical optimization methods.

[0004] The slime mold algorithm is a metaheuristic algorithm based on the foraging behavior and morphological changes of natural slime molds. Faced with complex dynamic environments, multiple constraints, and high-dimensional search spaces, traditional path planning algorithms lack computational efficiency, global search capabilities, and adaptability. However, the slime mold algorithm, with its simple principle, stability, and scalability, has been increasingly adopted in the field of path planning. However, the slime mold algorithm also suffers from issues such as reliance on initial solutions, low convergence accuracy, and a tendency to fall into local optima. Furthermore, the paths it plans are often poorly smooth and contain unnecessary turning points, increasing safety risks for unmanned agricultural machinery during operation. Summary of the Invention

[0005] In view of this, the present invention provides a real-time obstacle avoidance trajectory planning method, equipment and medium for unmanned agricultural machinery, which realizes real-time obstacle avoidance trajectory planning for unmanned agricultural machinery based on the improved slime mold algorithm based on multi-strategy fusion, and is used to obtain a smooth, stable, safe and collision-free optimal trajectory that meets the kinematic and dynamic constraints of agricultural machinery in real time in a complex dynamic environment.

[0006] To achieve the above purpose, the technical solution of the present invention is as follows:

[0007] A real-time obstacle avoidance trajectory planning method for unmanned agricultural machinery includes the following steps: constructing a local driving environment in a Frenet coordinate system based on acquired road environment information; constructing an objective function based on constraints such as smoothness, safety, vehicle dynamics, and vehicle kinematics; adopting a spherical vector-based encoding mechanism and a behavior-decision-oriented search agent initialization strategy to set the distribution position of a search agent group in the initial stage; introducing a staged target optimization strategy to determine the optimal fitness and the worst fitness of the group; adopting a staged dynamic update mechanism and a dynamic nonlinear reduction strategy to update individual positions; and determining the optimal planning result based on an adaptive iterative truncation mechanism to ultimately obtain a trajectory result with spatial and temporal information.

[0008] The steps include:

[0009] Step 1: Obtain the width and centerline of the mechanized farming path, the positioning of the agricultural machinery, and the location, outline, and speed of obstacles;

[0010] Step 2: Construct the local driving environment of the mechanized farming road in the Frenet coordinate system and transform the path planning problem into a search problem in the SL graph;

[0011] Step 3: Construct the objective function F based on smoothness, safety, vehicle dynamics and vehicle kinematics constraints safe and F smooth , where F safe Considering security constraints, F smooth Consider smoothness and stability;

[0012] Step 4: Using a spherical vector-based encoding mechanism and combining it with a behavioral decision-oriented initialization strategy, the distribution position of the search agent group in the initial stage is set;

[0013] Step 5: Introduce a phased target optimization strategy to sort the fitness of the search agent group in ascending order and determine the optimal fitness and the worst fitness;

[0014] Step 6: Use a phased dynamic update mechanism and a dynamic nonlinear decreasing strategy to update the search agent position;

[0015] Step 7: Determine the optimal planning result based on the adaptive iterative truncation mechanism;

[0016] Step 8: Construct an ST graph based on the optimal path planning results. For dynamic obstacles, calculate the projected area they occupy in the ST graph based on their state variables, thereby transforming the speed planning problem into a search problem in the ST graph.

[0017] Step 9: Determine the optimal speed planning result by using staged target optimization, staged dynamic update, dynamic nonlinear reduction and adaptive iterative truncation mechanism;

[0018] Step 10: Couple the path planning and velocity planning results to finally obtain a trajectory result with spatial and temporal information.

[0019] The specific implementation of step 4 is as follows:

[0020] Step 4.1: Determine the size of the search agent population popsize, the dimension of each search agent n, the maximum number of algorithm iterations max_it, the lower bound LB of the search space, and the upper bound UB of the search space;

[0021] Step 4.2: Using the path encoding mechanism based on spherical vectors, let X i is the i-th candidate solution in the group, specifically expressed as X i =(θ i1 ,θ i2 ,...,θ in ), and its corresponding P i =[(s i0 ,l i0 ),(s i1 ,l i1 ),(s i2 ,l i2 ),...,(s in ,l in )], where (s0,l0) is the position coordinate of the planning starting point;

[0022] Search by ProxyX i To path P i The mapping is performed according to the following formula:

[0023]

[0024] Where ds represents the longitudinal displacement between adjacent path points, θ∈[-π,π], j represents the jth dimension, j=1.2.3…n;

[0025] Step 4.3: Combined with the initialization strategy based on behavioral decision-making, the agricultural machinery divides the detour methods for each obstacle into left detour and right detour during path planning. The initialization formula for the search agent group is determined as follows:

[0026]

[0027] Where rand∈[0,1], dimension is n, and MB=0.

[0028] The specific implementation of step 5 is as follows:

[0029] Step 5.1: Decompose the entire algorithm into two stages. The first stage is to quickly screen collision-free paths in the search space and calculate the F of the search agent. safe Sort and determine the best fitness and the worst fitness;

[0030] Step 5.2: When the number of collision-free paths N safe When the number of iterations exceeds 2 / 3 of the group size or the current number of iterations t exceeds 1 / 2 of the maximum number of iterations, all current position information is retained and the second stage is entered;

[0031] Step 5.3: In Phase 2, based on F smooth Determine the best and worst fitness.

[0032] The specific implementation of step 6 is as follows:

[0033] Step 6.1: Update the path using a phased dynamic update mechanism:

[0034]

[0035] Where, represents the average position of the search agents in the first half after fitness sorting;

[0036] For different stages, temp is defined as follows:

[0037]

[0038] Where obs_l represents the horizontal coordinate of the obstacle; obs_judge represents whether the path collides with the obstacle. If so, obs_judge = 1; otherwise, obs_judge = 0; Stage_1 represents stage 1, and Stage_2 represents stage 2.

[0039] Step 6.2: The dynamic nonlinear decreasing strategy is expressed as follows:

[0040]

[0041] Where, vb max and VB min They are The maximum and minimum values of w are as follows, α is the adjustment coefficient; max and w min are the maximum and minimum values of the weight factor w, respectively, and β is the adjustment coefficient.

[0042] The specific implementation of step 7 is as follows: during the iterative process of phase 2, the standard deviation of the optimal fitness is calculated. If the standard deviation is less than the set constant δ and the current optimal path has no collision, the iterative process is automatically terminated.

[0043] The present invention also proposes an electronic device, which includes a processor and a memory for storing executable instructions of the processor; the processor is used to read the executable instructions from the memory and execute the instructions to implement the real-time obstacle avoidance trajectory planning method for unmanned agricultural machinery described in the present invention.

[0044] The present invention also proposes a computer-readable storage medium, which stores a computer program. The computer program is used to execute the real-time obstacle avoidance trajectory planning method for unmanned agricultural machinery described in the present invention.

[0045] Beneficial effects:

[0046] 1. By adopting a spherical vector-based path encoding mechanism combined with an initialization strategy based on behavioral decision-making, the method of the present invention can narrow the search space, better control the vehicle steering angle, and more efficiently obtain a trajectory that meets the vehicle's kinematic and dynamic constraints, thereby improving the efficiency and accuracy of downstream control tasks.

[0047] 2. The method of the present invention introduces a phased target optimization strategy and a phased dynamic update mechanism, decomposing the entire slime mold algorithm into two stages. The first stage quickly screens safe and collision-free paths in the search space, and the second stage smoothes the collision-free paths. These targeted position updates can effectively avoid the mutual interference between safety and smoothness during trajectory planning.

[0048] 3. The method of the present invention adopts a dynamic nonlinear decreasing strategy, which is beneficial for the algorithm to get rid of local extreme values and conduct global search in the early stage, avoiding falling into local optimality, while improving the search of the solution space more finely in the later stage, balancing the global and local search capabilities, and improving the convergence accuracy and speed.

[0049] 4. In the method of the present invention, the adaptive iterative truncation mechanism avoids unnecessary iterations and resource waste, improves the applicability of the algorithm in real-time planning, and improves the overall performance of the algorithm without sacrificing the solution quality.

[0050] 5. The device of the present invention is used to implement the method of the present invention. By adopting a path encoding mechanism based on spherical vectors and combining it with an initialization strategy based on behavioral decision-making, it can narrow the search space, better control the vehicle steering angle, and more efficiently obtain a trajectory that meets the vehicle's kinematic and dynamic constraints, thereby improving the efficiency and accuracy of downstream control tasks.

[0051] 6. The medium of the present invention is used to implement the method of the present invention. By adopting a path encoding mechanism based on spherical vectors and combining it with an initialization strategy based on behavioral decision-making, it can narrow the search space, better control the vehicle steering angle, and more efficiently obtain a trajectory that meets the vehicle's kinematic and dynamic constraints, thereby improving the efficiency and accuracy of downstream control tasks. BRIEF DESCRIPTION OF THE DRAWINGS

[0052] Figure 1 This is a flowchart of real-time obstacle avoidance trajectory planning for unmanned agricultural machinery using the improved slime mold algorithm according to an embodiment of the present invention.

[0053] Figure 2 This is a flowchart of the multi-strategy fusion improved slime mold algorithm according to an embodiment of the present invention.

[0054] Figure 3 Schematic diagram of Cartesian coordinates and Frenet coordinate systems according to an embodiment of the present invention.

[0055] Figure 4 It is the path planning SL graph of an embodiment of the present invention.

[0056] Figure 5 It is a speed planning ST diagram of an embodiment of the present invention.

[0057] Figure 6 A schematic structural diagram of an electronic device provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0058] The present invention is described in detail below with reference to the accompanying drawings and embodiments.

[0059] The flowchart of the real-time obstacle avoidance trajectory planning of unmanned agricultural machinery using the improved slime mold algorithm according to the embodiment of the present invention is as follows: Figure 1 As shown, the following steps are included:

[0060] Step 1: Use the map to extract the width and centerline information of the mechanized farming path. Use the lidar to detect the position, outline, and speed of obstacles in real time, and rely on integrated navigation to obtain the positioning information of the agricultural machinery in real time. The obstacles include static obstacles and dynamic obstacles. Among them, before the agricultural machinery performs trajectory planning, it is necessary to obtain information such as the width of the mechanized farming path, the centerline of the mechanized farming path, the positioning of the agricultural machinery, and the position, outline, and speed of the obstacles.

[0061] Step 2: To cope with the different curvatures of mechanized farming roads and more accurately describe the relationship between the positions of agricultural machinery, obstacles, and mechanized farming roads, a local driving environment of the mechanized farming roads is constructed in the Frenet coordinate system, transforming the path planning problem into a path search problem in the SL graph. Figure 3The comparison between the Cartesian coordinate system and the Frenet coordinate system is demonstrated, and the local driving environment of the mechanized farming road in the Frenet coordinate system is constructed. The path planning problem is transformed into a path search problem in the SL graph, and the speed planning problem is transformed into a search problem in the ST graph.

[0062] Step 3: Construct the objective function F based on the constraints of smoothness, safety, vehicle dynamics and vehicle kinematics. safe and F smooth , where F safe Mainly consider safety constraints, which can be expressed by the distance between the point and the obstacle; F smooth It focuses on smoothness and stability, which can be indirectly deduced through the geometric relationship of discrete points.

[0063] Step 4: Using a spherical vector-based encoding mechanism and combining it with a behavioral decision-oriented initialization strategy, the distribution position of the search agent group in the initial stage is set;

[0064] Step 5: Introduce a phased target optimization strategy to sort the fitness of the search agent group in ascending order and determine the optimal fitness and the worst fitness;

[0065] Step 6: Update the search agent position using a phased dynamic update mechanism and a dynamic nonlinear decreasing strategy;

[0066] Step 7: Determine the optimal path planning result based on the adaptive iterative truncation mechanism;

[0067] Step 8: Based on the optimal path planning results, the ST graph is constructed. For dynamic obstacles, the projected area occupied by them in the ST graph is calculated according to their state quantities, thereby converting the speed planning problem into a search problem in the ST graph.

[0068] Step 9, referring to the path planning process in the SL graph, adopts a path encoding mechanism based on spherical vectors, and uses strategies such as staged target optimization, staged dynamic update, dynamic nonlinear reduction, and adaptive iterative truncation mechanism to determine the optimal speed planning result;

[0069] Step 10: Couple the path planning and velocity planning results to finally obtain a trajectory result with spatial and temporal information.

[0070] Figure 4 is the path planning SL graph of an embodiment of the present invention, Figure 5 It is the speed planning ST graph of the embodiment of the present invention. Preferably, a path encoding mechanism based on spherical vectors is adopted, and an initialization strategy guided by behavioral decision-making is combined to set the distribution position of the search agent group in the initial stage, including the following steps:

[0071] Determine the size of the search agent population popsize, the dimension of each search agent n, the maximum number of algorithm iterations max_it, the lower bound LB of the search space, and the upper bound UB of the search space;

[0072] Using the path encoding mechanism based on spherical vectors, let X i is the i-th candidate solution in the group, expressed as follows: X i =(θ i1 ,θ i2 ,...,θ in ), where θ∈[-π, π]. with X i The corresponding path P i =[(s i1 , l i2 )(s i1 , l i1 ), (s i2 , l i2 ),...,(s in , l in )], where (s0, l0) is the location coordinate of the planning starting point. i To path P i The mapping is performed according to the following formula:

[0073]

[0074] Where ds represents the longitudinal displacement between adjacent path points; j represents the jth dimension, j = 1.2.3...n;

[0075] Combined with the initialization strategy based on behavioral decision-making, the agricultural machinery's detour methods for each obstacle are divided into "left detour" and "right detour". The initialization formula for the search agent group is determined as follows:

[0076]

[0077] Where rand∈[0,1], dimension is n, and MB=0.

[0078] Preferably, a phased target optimization strategy is introduced to sort the fitness of the search agent group in ascending order, including the following steps:

[0079] The entire slime mold algorithm is decomposed into two stages. Figure 2 The flowchart of the improved slime mold algorithm is shown as follows:

[0080] Phase 1 quickly screens collision-free paths in the search space and calculates the F of the search agent. safe And sort them to determine the best fitness and the worst fitness; when the number of collision-free paths N safeWhen the number of iterations exceeds 2 / 3 of the group size or the current number of iterations t exceeds 1 / 2 of the maximum number of iterations, all current position information is retained and the second stage is entered. The main purpose of the second stage is to obtain a smoother collision-free path that meets kinematic constraints. smooth Determine the optimal fitness value and the worst fitness.

[0081] Preferably, a phased dynamic update mechanism and a dynamic nonlinear decreasing strategy are used to update the path in the following specific implementations:

[0082]

[0083] in, Represents the average position of the search agents in the first half after fitness sorting. vb max and VB min They are The maximum and minimum values of w are as follows, α is the adjustment coefficient; max and w min are the maximum and minimum values of the weight factor w, respectively, and β is the adjustment coefficient;

[0084] For different stages, temp is defined as follows:

[0085]

[0086] Where obs_l represents the horizontal coordinate of the obstacle. obs_judge represents whether the path collides with the obstacle. If it collides, obs_judge = 1, otherwise, obs_judge = 0. Stage_1 represents stage 1, and Stage_2 represents stage 2.

[0087] Preferably, the adaptive iterative truncation mechanism is specifically as follows: during the iterative process of the second stage of the algorithm, the standard deviation of the optimal fitness is calculated; if the standard deviation is less than a set constant δ and the current optimal path has no collision, the iterative process is automatically terminated.

[0088] The embodiment of the present application also provides an electronic device, Figure 6The structure of an electronic device provided by an embodiment of the present invention is shown. For example, the electronic device 60 may include a processor 61, a memory 62, and a transmission device 63. The processor 61 is used to execute the real-time obstacle avoidance trajectory planning method, device, and medium for unmanned agricultural machinery mentioned in the above-mentioned embodiments. The processor and memory may be connected via a bus or other means, with bus connection being used as an example. The transmission device may be connected to the processor and memory via a wired or wireless connection. The memory, as a non-transitory computer-readable storage medium, may be used to store non-transitory software programs, non-transitory computer executable programs, and modules, such as the program instructions / modules corresponding to the real-time obstacle avoidance trajectory planning method, device, and medium for unmanned agricultural machinery in the embodiments of the present application. The processor executes the non-transitory software programs, instructions, and modules stored in the memory to execute various functional applications and data processing of the processor, thereby implementing the real-time obstacle avoidance trajectory planning method, device, and medium for unmanned agricultural machinery in the above-mentioned method embodiments. The memory may include a program storage area and a data storage area. The program storage area may store an operating system and application programs required for at least one function; the data storage area may store data created by the processor, etc. In addition, the memory may include a high-speed random access memory and may also include a non-volatile memory, such as at least one disk storage device, a flash memory device, or other non-volatile solid-state storage device. In some embodiments, the memory may optionally include a memory remotely located relative to the processor, and these remote memories may be connected to the processor via a network. Examples of the above-mentioned networks include but are not limited to the Internet, an intranet, a local area network, a mobile communication network, and a combination thereof. The one or more modules are stored in the memory, and when executed by the processor, the unmanned agricultural machinery real-time obstacle avoidance trajectory planning method, device, and medium in the embodiment are executed.

[0089] As another aspect, the present application also provides a computer-readable storage medium, which may be the computer-readable storage medium included in the apparatus described in the above embodiment; or it may be a computer-readable storage medium that exists independently and is not assembled into the device. The computer-readable storage medium may be a tangible storage medium, such as a random access memory (RAM), internal memory, read-only memory (ROM), electrically programmable ROM, electrically erasable programmable ROM, register, floppy disk, hard disk, removable storage disk, CD-ROM, or any other form of storage medium known in the technical field. The computer-readable storage medium stores one or more programs, which are used by one or more processors to execute the unmanned agricultural machinery real-time obstacle avoidance trajectory planning method, device, and medium described in the present application.

[0090] In summary, the above are only preferred embodiments of the present invention and are not intended to limit the scope of protection of the present invention. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A real-time obstacle avoidance trajectory planning method for unmanned agricultural machinery, characterized in that: The following steps are involved: Based on the acquired road environment information, a local driving environment in the Frenet coordinate system is constructed. An objective function is constructed based on constraints such as smoothness, safety, vehicle dynamics, and vehicle kinematics. A spherical vector-based encoding mechanism and a behavior-based decision-making-oriented search agent initialization strategy are used to set the distribution position of the search agent group in the initial stage. A staged target optimization strategy is introduced to determine the optimal and worst fitness of the group. An in-stage dynamic update mechanism and a dynamic nonlinear decrement strategy are used to update individual positions. Based on the adaptive iterative truncation mechanism, the optimal planning result is determined, and finally a trajectory result with spatial and temporal information is obtained.

2. The method according to claim 1, wherein The following steps are involved: Step 1: Obtain the width and centerline of the mechanized farming path, the positioning of the agricultural machinery, and the location, outline, and speed of obstacles; Step 2: Construct the local driving environment of the mechanized farming road in the Frenet coordinate system and transform the path planning problem into a search problem in the SL graph; Step 3: Construct the objective function F based on smoothness, safety, vehicle dynamics and vehicle kinematics constraints safe and F smooth , where F safe Considering security constraints, F smooth Consider smoothness and stability; Step 4: Using a spherical vector-based encoding mechanism and combining it with a behavioral decision-oriented initialization strategy, the distribution position of the search agent group in the initial stage is set; Step 5: Introduce a phased target optimization strategy to sort the fitness of the search agent group in ascending order and determine the optimal fitness and the worst fitness; Step 6: Use a phased dynamic update mechanism and a dynamic nonlinear decreasing strategy to update the search agent position; Step 7: Determine the optimal planning result based on the adaptive iterative truncation mechanism; Step 8: Construct an ST graph based on the optimal path planning results. For dynamic obstacles, calculate the projected area they occupy in the ST graph based on their state variables, thereby transforming the speed planning problem into a search problem in the ST graph. Step 9: Determine the optimal speed planning result by using staged target optimization, staged dynamic update, dynamic nonlinear reduction and adaptive iterative truncation mechanism; Step 10: Couple the path planning and velocity planning results to finally obtain a trajectory result with spatial and temporal information.

3. The method according to claim 2, wherein The specific implementation of step 4 is as follows: Step 4.1: Determine the size of the search agent population popsize, the dimension of each search agent n, the maximum number of algorithm iterations max_it, the lower bound LB of the search space, and the upper bound UB of the search space; Step 4.2: Using the path encoding mechanism based on spherical vectors, let X i is the i-th candidate solution in the group, specifically expressed as X i =(θ i1 ,θ i2 ,...,θ in ), and its corresponding P i =[(s i0 , l i0 ), (s i1 , l i1 ), (s i2 , l i2 ),...,(s in , l in )], where (s0, l0) is the position coordinate of the planning starting point; Search by ProxyX i To path P i The mapping is performed according to the following formula: Where ds represents the longitudinal displacement between adjacent path points, θ∈[-π,π], j represents the jth dimension, j=1.2.3…n; Step 4.3: Combined with the initialization strategy based on behavioral decision-making, the agricultural machinery divides the detour methods for each obstacle into left detour and right detour during path planning. The initialization formula for the search agent group is determined as follows: Where rand∈[0,1], dimension is n, and MB=0.

4. The method according to claim 2 or 3, wherein: The specific implementation of step 5 is as follows: Step 5.1: The entire algorithm is decomposed into two stages. In the first stage, the collision-free path is quickly screened in the search space, and the F of the search agent is safe Sort and determine the best fitness and the worst fitness; Step 5.2: When the number of collision-free paths N safe When the number of iterations exceeds 2 / 3 of the group size or the current number of iterations t exceeds 1 / 2 of the maximum number of iterations, all current position information is retained and the second stage is entered; Step 5.3: In Phase 2, based on F smooth Determine the best and worst fitness.

5. The method according to claim 4, wherein The specific implementation of step 6 is as follows: Step 6.1: Update the path using a phased dynamic update mechanism: Where, represents the average position of the search agents in the first half after fitness sorting; For different stages, temp is defined as follows: Where obs_l represents the horizontal coordinate of the obstacle; obs_judge represents whether the path collides with the obstacle. If so, obs_judge = 1; otherwise, obs_judge = 0; Stage_1 represents stage 1, and Stage_2 represents stage 2. Step 6.2: The dynamic nonlinear decreasing strategy is expressed as follows: Where, vb max and VB min They are The maximum and minimum values of , α is the adjustment coefficient; w max and w min are the maximum and minimum values of the weight factor w, respectively, and β is the adjustment coefficient.

6. The method according to claim 2 or 3, wherein: The specific implementation of step 7 is as follows: during the iterative process of stage 2, the standard deviation of the optimal fitness is calculated. If the standard deviation is less than the set constant δ and the current optimal path has no collision, the iterative process is automatically terminated.

7. An electronic device, characterized in that: The electronic device includes a processor and a memory for storing executable instructions of the processor; the processor is used to read the executable instructions from the memory and execute the instructions to implement the real-time obstacle avoidance trajectory planning method for unmanned agricultural machinery as described in any one of claims 1-6 above.

8. A computer-readable storage medium, characterized in that The storage medium stores a computer program, which is used to execute the real-time obstacle avoidance trajectory planning method for unmanned agricultural machinery as described in any one of claims 1 to 6.

Images