Intelligent navigation system and method for agricultural machinery

Abstract

The invention discloses an intelligent navigation system and method for agricultural machinery, and relates to the technical field of vehicle control, and the system comprises a sensing layer which obtains environment sensing data; the decision-making layer is used for determining obstacle information according to the image in front of the vehicle and the three-dimensional point cloud information by adopting an edge computing platform; performing path planning according to the obstacle information and the agricultural machine position information; performing path tracking decision control by adopting an improved pure tracking algorithm, and calculating a target steering angle of a steering wheel of the agricultural machine; the improved pure tracking algorithm adopts the vehicle speed, the path curvature and the tracking error to calculate the pre-anchor point distance; the control layer converts a decision result of the decision layer into a control instruction by using a single chip microcomputer; the decision result comprises a path planning result and a target steering angle of an agricultural machine steering wheel; and the execution layer receives the control instruction of the control layer and controls the agricultural machinery to execute the action corresponding to the control instruction, so that the flexibility, expansibility and reliability of the system are improved, the size is small, the development cost is low, and the control precision and response speed of path tracking are improved.

Description

Technical Field

[0001] The present application relates to the field of vehicle control technology, and in particular to an intelligent navigation system and method for agricultural machinery. Background Art

[0002] The automatic navigation technology of agricultural machinery vehicles is a key technology to realize the unmanned precision operation of agricultural machinery. Due to the limitations of the computing power of the computing platform or the limitations of the hardware platform structure design, the current agricultural machinery automatic navigation system is difficult to process a large amount of data from sensors such as images and LiDAR, so it is difficult to be compatible with the function of real-time obstacle detection. In addition, most of the current agricultural machinery automatic navigation systems are mainly designed based on an integrated computing platform. Since all data processing and control algorithms are concentrated on one platform, the functional expansion and upgrade of the system are limited, lacking flexibility and scalability. If the computing platform fails, the function of the entire system may be affected, and the reliability of the system is low. In order to improve the flexibility and scalability of the automatic navigation system. Previously, the team developed an agricultural machinery automatic navigation system based on industrial computers and PLC controllers, and proposed a distributed control automatic navigation system. The industrial computer is responsible for processing complex decision-making algorithms such as path planning and path tracking algorithms, and the PLC controller is responsible for executing control algorithms such as steering and braking. However, the industrial computer and PLC controller used in this system are large in size and high in cost.

[0003] In addition, current research on agricultural machinery path tracking control methods usually only focuses on how to accurately calculate the front wheel steering angle based on the path deviation, and often ignores how to quickly and effectively adjust the actual front wheel steering angle of the agricultural machinery after the front wheel steering angle is calculated. Especially in complex field environments, if the steering adjustment is not fast enough, even if the accurate front wheel steering angle is obtained in time, the path tracking accuracy may still be low. Therefore, after accurately calculating the front wheel steering angle of the agricultural machinery based on the agricultural machinery path tracking method, how to quickly and accurately adjust the direction of the agricultural machinery has become another key problem in current research. Summary of the invention

[0004] The purpose of this application is to provide an intelligent navigation system and method for agricultural machinery, which can improve the flexibility, scalability and reliability of the system, have a small size, low development cost, and improve the control accuracy and response speed of the agricultural machinery path tracking algorithm in complex environments and multiple paths.

[0005] To achieve the above objectives, this application provides the following solutions.

[0006] In a first aspect, the present application provides an agricultural machinery intelligent navigation system, comprising: The perception layer is used to obtain environmental perception data; the environmental perception data includes agricultural machinery position information, vehicle front image, three-dimensional point cloud information and vehicle posture information; The decision layer is used to: use an edge computing platform to determine obstacle information according to the image in front of the vehicle and the three-dimensional point cloud information; the obstacle information includes obstacle category and obstacle distance; perform path planning according to the obstacle information and the agricultural machinery position information to obtain a path planning result; use an improved pure tracking algorithm to perform path tracking decision control according to the path planning result, and calculate the target steering angle of the steering wheel of the agricultural machinery; the improved pure tracking algorithm uses vehicle speed, path curvature and tracking error to calculate the pre-anchor point distance; The control layer is used to: use a single-chip microcomputer to convert the decision results of the decision layer into control instructions; the decision results include path planning results and the target steering angle of the agricultural machinery steering wheel; The execution layer is used to receive control instructions from the control layer and control the agricultural machinery to execute actions corresponding to the control instructions.

[0007] Optionally, the execution layer is further used to collect the actual front wheel steering angle of the agricultural machine, and feed the actual front wheel steering angle back to the control layer; The control layer is used to: utilize a single chip microcomputer and a fuzzy PID control algorithm to perform steering control according to the actual front wheel steering angle and the target steering angle to achieve closed-loop control.

[0008] Optionally, the calculation formula of the pre-anchor point distance is as follows: ; in, represents the pre-anchor distance, is the speed of the agricultural machinery, is the path tracking error, is the path curvature, are the vehicle speed coefficient, path tracking error coefficient and path curvature coefficient respectively, is a constant term.

[0009] Optionally, in adopting the improved pure tracking algorithm, performing path tracking decision control according to the path planning result, and calculating the target steering angle of the agricultural machinery steering wheel, the decision layer is used to: The polynomial coefficients are optimized by using an ant colony algorithm to obtain optimal polynomial coefficients; the polynomial coefficients include a vehicle speed coefficient, a path tracking error coefficient, and a path curvature coefficient; Calculating the pre-anchor point distance according to the optimal polynomial coefficients; According to the pre-anchor point distance, a target steering angle of the agricultural machinery steering wheel is calculated.

[0010] Optionally, the calculation formula of the target steering angle of the agricultural machinery steering wheel is as follows: ; In the formula, for The target steering angle of the agricultural machinery steering wheel at any moment, is the vehicle wheelbase, for The angle between the pre-anchor point and the vehicle body at all times.

[0011] In a second aspect, the present application provides an agricultural machinery intelligent navigation method based on the above-mentioned agricultural machinery intelligent navigation system, comprising: Acquire environmental perception data; the environmental perception data includes agricultural machinery position information, vehicle front image, three-dimensional point cloud information and vehicle posture information; An edge computing platform is used to determine obstacle information according to the image in front of the vehicle and the three-dimensional point cloud information; the obstacle information includes obstacle category and obstacle distance; path planning is performed according to the obstacle information and the agricultural machinery position information to obtain a path planning result; an improved pure tracking algorithm is used to perform path tracking decision control according to the path planning result, and a target steering angle of the steering wheel of the agricultural machinery is calculated; the improved pure tracking algorithm uses vehicle speed, path curvature and tracking error to calculate the pre-anchor point distance; The single chip microcomputer is used to convert the decision results of the decision layer into control instructions; the decision results include the path planning results and the target steering angle of the agricultural machinery steering wheel; Receive control instructions from the control layer and control the agricultural machinery to execute actions corresponding to the control instructions.

[0012] Optionally, the agricultural machinery intelligent navigation method further includes: Collecting the actual front wheel steering angle of the agricultural machinery and feeding back the actual front wheel steering angle to the control layer; By using a single chip microcomputer and a fuzzy PID control algorithm, steering control is performed according to the actual front wheel steering angle and the target steering angle to achieve closed-loop control.

[0013] Optionally, the calculation formula of the pre-anchor point distance is as follows: ; in, represents the pre-anchor distance, is the speed of the agricultural machinery, is the path tracking error, is the path curvature, are the vehicle speed coefficient, path tracking error coefficient and path curvature coefficient respectively, is a constant term.

[0014] Optionally, an improved pure tracking algorithm is used to perform path tracking decision control according to the path planning result, and a target steering angle of the steering wheel of the agricultural machine is calculated, specifically including: The polynomial coefficients are optimized by using an ant colony algorithm to obtain optimal polynomial coefficients; the polynomial coefficients include a vehicle speed coefficient, a path tracking error coefficient, and a path curvature coefficient; Calculating the pre-anchor point distance according to the optimal polynomial coefficients; According to the pre-anchor point distance, a target steering angle of the agricultural machinery steering wheel is calculated.

[0015] Optionally, the calculation formula of the target steering angle of the agricultural machinery steering wheel is as follows: ; In the formula, for The target steering angle of the agricultural machinery steering wheel at any moment, is the vehicle wheelbase, for The angle between the pre-anchor point and the vehicle body at all times.

[0016] According to the specific embodiments provided by the present application, the present application discloses the following technical effects: The present application provides an intelligent navigation system and method for agricultural machinery, which is compatible with both environmental perception and automatic navigation functions. The edge computing platform has powerful computing power and is responsible for processing complex environmental perception and navigation algorithms. The single chip microcomputer (MCU) is responsible for executing agricultural machinery motion control tasks to ensure rapid response to navigation instructions from the edge platform. The system has the following advantages: (1) The intelligent navigation tasks are distributed and the system can reasonably and effectively utilize computing resources, while processing complex environmental perception and automatic navigation tasks, and is compatible with environmental perception and automatic navigation functions; (2) The modular design of the loosely coupled system makes it more flexible in system upgrades and maintenance, and it is small in size and has low development costs; (3) In the event of a failure of the edge computing platform, the MCU can continue to run basic control tasks to a certain extent, improving the reliability and stability of the system. The improved pure tracking algorithm comprehensively considers multiple factors such as vehicle speed, path curvature and path tracking error, and uses a polynomial function of speed, path curvature and tracking error to dynamically calculate the pre-anchor point distance in the pure tracking algorithm. The improved pure tracking algorithm can quickly and accurately calculate the front wheel steering angle, and further uses a fuzzy PID control algorithm to control the rotation of the electric steering wheel, thereby realizing rapid and effective adjustment of the front wheel steering angle, and improving the control accuracy and response speed of path tracking. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0018] Figure 1 A framework diagram of an intelligent navigation system for agricultural machinery provided in accordance with one embodiment of the present application.

[0019] Figure 2 A schematic diagram of the construction and installation of an intelligent navigation hardware platform for agricultural machinery provided in one embodiment of the present application.

[0020] Figure 3 A schematic diagram of a pure tracking algorithm provided in an embodiment of the present application.

[0021] Figure 4 A schematic diagram of a polynomial coefficient optimization process provided in one embodiment of the present application.

[0022] Figure 5 A schematic diagram of the fuzzy PID control principle provided in one embodiment of the present application.

[0023] Figure 6 A flowchart of an intelligent navigation method for agricultural machinery provided in one embodiment of the present application. DETAILED DESCRIPTION

[0024] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0025] In order to make the above-mentioned objects, features and advantages of the present application more obvious and easy to understand, the present application is further described in detail below with reference to the accompanying drawings and specific implementation methods.

[0026] In an exemplary embodiment, Figure 1 As shown, an agricultural machinery intelligent navigation system is provided, including a perception layer, a decision layer, a control layer and an execution layer.

[0027] The perception layer is used to obtain environmental perception data; the environmental perception data includes agricultural machinery position information, vehicle front image, three-dimensional point cloud information and vehicle posture information.

[0028] The perception layer is mainly responsible for acquiring a variety of environmental perception data, such as data collected by sensors such as the Global Navigation Satellite System (GNSS), RGB cameras, 3D LiDAR, and Inertial Measurement Unit (IMU), providing basic information for subsequent intelligent navigation decisions.

[0029] GNSS sensors provide agricultural machinery location information. This application uses network RTK (Real-Time Ki nematic technology for high-precision positioning. The RGB camera is used to collect image information in front of the vehicle, and the JIERUIXINGTONG HF500 camera can be used. The 3D LiDAR is used to obtain rich 3D point cloud information around, and the Velodyne HDL-32E 3D LiDAR can be used. The IMU sensor provides vehicle posture information. GNSS and IMU information are sent to the edge computing platform through the serial port, and the image in front of the vehicle and the 3D point cloud information are sent to the edge computing platform through USB and network cable (Net cable) respectively. The GNSS sensor and IMU sensor communicate with the edge computing platform through TTL, the RGB camera communicates with the edge computing platform through the USB data cable, and the 3D LiDAR communicates with the edge computing platform through the Net cable.

[0030] In order to improve the control accuracy and response speed of the agricultural machinery path tracking algorithm under complex environments and multiple paths, this application proposes a dual closed-loop path tracking control method based on an improved pure tracking model and fuzzy PID, using an improved pure tracking algorithm for decision control, and outputting an accurate steering angle according to the tracking deviation. After obtaining the accurate front wheel steering angle, the fuzzy PID algorithm is used for steering control, and the steering of the agricultural machinery is adjusted quickly and stably according to the steering angle. The polynomial function of this application comprehensively considers multiple factors such as vehicle speed, path curvature and path tracking error, and uses a polynomial function of speed, path curvature and tracking error to dynamically calculate the pre-anchor point distance in the pure tracking algorithm, and uses an ant colony algorithm to optimize the coefficients of the polynomial function, so that the system can flexibly adapt to different vehicle speeds and path complexities (path curvature and path tracking error changes). The improved pure tracking algorithm can quickly and accurately calculate the front wheel steering angle. After the front wheel steering angle is accurately obtained by using the improved pure tracking algorithm, the fuzzy PID control algorithm is further used to control the rotation of the electric steering wheel, thereby realizing the rapid and effective adjustment of the front wheel steering angle. Fuzzy PID control combines the advantages of fuzzy logic and traditional PID control. It can achieve rapid response and precise control of the steering wheel by adaptively adjusting control parameters, ensuring that agricultural machinery can always operate efficiently and stably along the predetermined path in high-precision agricultural tasks. It can adaptively adjust control parameters in complex and changing environments, significantly improving the response speed and accuracy of the system.

[0031] The decision layer is used to: use an edge computing platform to determine obstacle information based on the image in front of the vehicle and the three-dimensional point cloud information; the obstacle information includes obstacle category and obstacle distance; perform path planning based on the obstacle information and the agricultural machinery position information to obtain a path planning result; use an improved pure tracking algorithm to perform path tracking decision control based on the path planning result, and calculate the target steering angle of the agricultural machinery steering wheel; the improved pure tracking algorithm uses vehicle speed, path curvature and tracking error to calculate the pre-anchor point distance.

[0032] This application realizes obstacle detection based on the fusion of RGB camera and 3D LiDAR. The YOLO v5 algorithm is used to identify the category of obstacles based on the image in front of the vehicle collected by the RGB camera. The distance information of obstacles is determined using the three-dimensional point cloud information collected by 3D LiDAR. Finally, the three-dimensional point cloud information is projected into the image in front of the vehicle, and the category and distance information of obstacles in the image in front of the vehicle are obtained at the same time.

[0033] When the agricultural machinery is driven manually, the longitude and latitude position information of the precise agricultural machinery (i.e. the agricultural machinery position information) can be obtained in real time through environmental perception and positioning data acquisition. During the driving process of the agricultural machinery, the driving path is recorded and saved locally.

[0034] This application uses an improved pure tracking algorithm for path tracking control of agricultural machinery. The selection of the pre-anchor point distance is crucial in the pure tracking algorithm because it directly affects the accuracy of path tracking, the vehicle's driving stability, turning performance, and adaptability to different speeds. This application expresses the pre-anchor point distance as a polynomial function of vehicle speed, tracking error, and path curvature, as shown in the following formula.

[0035] (1).

[0036] in, represents the pre-anchor distance, is the speed of the agricultural machinery, is the path tracking error, is the path curvature, are the vehicle speed coefficient, path tracking error coefficient and path curvature coefficient respectively, is a constant term.

[0037] In adopting the improved pure tracking algorithm, performing path tracking decision control according to the path planning results, and calculating the target steering angle of the steering wheel of the agricultural machinery, the decision layer is used to: adopt the Ant Colony Algorithm (ACO) to optimize the polynomial coefficients to obtain the optimal polynomial coefficients; the polynomial coefficients include the vehicle speed coefficient, the path tracking error coefficient and the path curvature coefficient; calculate the pre-anchor point distance according to the optimal polynomial coefficients; and calculate the target steering angle of the steering wheel of the agricultural machinery according to the pre-anchor point distance.

[0038] The determination of the coefficients of the pre-anchor distance polynomial is the key to accurately describe the relationship between the pre-anchor distance and vehicle speed, error and curvature. This application uses an ant colony algorithm to optimize the polynomial coefficients so that the pre-anchor distance can dynamically adapt to changes in vehicle speed, path tracking error and path curvature.

[0039] Assuming the center of the vehicle's rear wheel A Can follow a certain turning radius R Drive to the pre-anchor point, and then according to the distance of the pre-anchor point , turning radius R The geometric relationship between the pre-anchor point and the vehicle body in the vehicle coordinates determines the front wheel steering angle . Figure 3 The meanings of the symbols are shown in Table 1.

[0040] Table 1 Figure 3 The meaning of each symbol

[0041] according to Figure 3 The geometric relationship can be obtained as shown in equation (2).

[0042] (2).

[0043] Simplifying the above formula (2) gives the following formula (3).

[0044] (3).

[0045] Based on the bicycle model, the relationship shown below can be obtained.

[0046] (4).

[0047] Combining equation (2) and equation (4), we get the calculation formula for the front wheel steering angle, as shown below.

[0048] (5).

[0049] Front wheel steering angle The angle between the pre-anchor point and the vehicle body A function of time t is obtained, so it is deduced from formula (5) that the target steering angle of the agricultural machinery steering wheel, that is, the calculation formula of the target front wheel steering angle is shown in the following formula.

[0050] (6).

[0051] In the formula, for The target steering angle of the agricultural machinery steering wheel at any moment, is the vehicle wheelbase, for The angle between the pre-anchor point and the vehicle body at all times.

[0052] ACO uses artificial ants to move in the problem space and uses pheromone and heuristic information to find the optimal solution to the problem. Ants release pheromones during foraging, and other ants are attracted by the pheromones and tend to choose paths with higher pheromone concentrations. This mechanism helps the ant colony find the shortest path. After the ants complete a path selection, the pheromone concentration is updated according to the quality of the path. The better the path, the more pheromones increase, and vice versa. The probability of an ant choosing a path is proportional to the visibility of the path and the pheromone concentration. The specific formula is shown below.

[0053] (7).

[0054] in, The ant starts from the path point To waypoint The probability of a path point represents a possible choice of any coefficient in the polynomial coefficients; It is Iterations pheromone concentration on; It is The distance heuristic function in the number of iterations; and They are the pheromone weight coefficient and the heuristic function weight coefficient respectively; allowed is the set of paths that the ants can choose, For the Iterations The pheromone concentration on the surface. Pheromones evaporate over time, preventing the algorithm from converging to the local optimum too early. The formula is as follows.

[0055] (8).

[0056] in, For the +1 iterations for edges The pheromone concentration on is the pheromone volatility coefficient, It's Ant The number of iterations in the edge The pheromone increment left on the path. The pheromone increment is usually related to the path length. The shorter the path, the greater the increment. The calculation formula of the pheromone increment is shown in the following formula.

[0057] (9).

[0058] (10).

[0059] in, is the number of ants, Ant No. The number of iterations in the edge The pheromone increment left on is the pheromone quantity constant, It's an ant The length of the path taken.

[0060] The fitness function design is a key step in the optimization process. This application defines the fitness function as the sum of path tracking errors. The fitness function is expressed as the following formula.

[0061] (11).

[0062] In the formula, is the fitness function, is the number of path points in the simulation, For the The path tracking error of each path point is calculated to ensure the adaptability of the polynomial coefficients and to fully optimize the polynomial coefficients to improve the path tracking performance of the vehicle under different paths, speeds and initial deviation conditions.

[0063] This application designs various types of simulation test paths, including straight segments, circular segments (turning radius from 4 meters to 10 meters, with intervals of 0.5 meters) and S-shaped curves (curvature changes from 4 meters to 10 meters, with intervals of 0.5 meters) to simulate driving scenarios under different turning conditions. The experiment also covers different vehicle speeds (0.5 m / s to 2 m / s, with intervals of 0.5 m / s) and initial deviations (0.5 m to 2 m, with intervals of 0.5 m) to fully simulate driving scenarios under different speeds and deviations. Through these diverse experiments, it is ensured that the path tracking method proposed in this application can achieve good path tracking effects under different speeds, path deviations and tracking path curvature conditions.

[0064] The process of optimizing polynomial coefficients based on ant colony algorithm is as follows: Figure 4 As shown in FIG, the method specifically includes: initializing the ant colony and randomly setting the polynomial coefficients; performing path tracking simulation to evaluate the tracking error. Then, the fitness value of the current path is calculated according to the fitness function. The fitness value of the current path is calculated by the path tracking error corresponding to all path points in the current iteration to summarize the error; judging whether the current fitness value meets the convergence condition, if so, outputting the best polynomial coefficients; if not, updating the pheromone, continuing the iteration, and gradually optimizing the polynomial coefficients through multiple simulations and adjustments to improve the path tracking performance.

[0065] The convergence condition is that the fitness value corresponding to the tracking error of the current path is less than 1 cm. In the above iterative process, when the tracking errors of all types of paths are less than 1 cm, the convergence condition is considered to be met. In the parameter optimization process of this application, the number of ants and the pheromone volatility coefficient , pheromone weight coefficient , heuristic function weight coefficient Total pheromone Q Set them to 20, 0.7, 1.5, 1.5 and 100 respectively.

[0066] The control layer is used to: use a single-chip microcomputer to convert the decision results of the decision layer into control instructions; the decision results include path planning results and the target steering angle of the agricultural machinery steering wheel.

[0067] The execution layer is used to receive control instructions from the control layer and control the agricultural machinery to execute actions corresponding to the control instructions.

[0068] The decision layer analyzes and makes decisions based on the data provided by the perception layer. The edge computing platform uses its powerful computing power to handle complex environmental perception and automatic navigation algorithms. In the Jetson Xavier NX edge computing platform, obstacles are detected based on the fusion method of RGB camera and 3D LiDAR. The YOLO v5 algorithm is used to detect the obstacle category, and the point cloud information is used to obtain the distance of the obstacle. This application uses an improved pure tracking algorithm for path tracking decision control, and calculates the steering wheel of the agricultural machinery ( Figure 2 The decision layer sends the intelligent navigation decision results to the control layer through the serial port, and the control layer converts the decision results of the decision layer into actual control instructions, such as direction, braking and other commands. The single-chip microcomputer can achieve fine control of the execution layer due to its high flexibility and strong real-time performance. The fuzzy PID control algorithm is used in the single-chip microcomputer for steering control to ensure that the front wheel steering angle is adjusted quickly and accurately. The execution layer receives the instructions from the control layer, executes the corresponding actions, and feeds back the actual execution status to the control layer to realize closed-loop control, ensuring that the intelligent navigation system drives along the predetermined path or avoids obstacles. The hardware composition and installation location of the intelligent navigation system for agricultural machinery proposed in this application are as follows: Figure 2 As shown in the figure, GNSS sensor, RGB camera, 3D LiDAR, IMU sensor, edge computing platform, angle sensor and single chip microcomputer are installed on the agricultural machinery vehicle, among which RGB camera and 3D LiDAR are installed on the front end of the agricultural machinery vehicle. The single chip microcomputer adopts STM32F4 single chip microcomputer. The edge computing platform can adopt Jetson Xavier NX edge computing platform. Figure 2 The display terminal is used to display the current driving status of the agricultural machinery vehicle, and the cylinder is a braking device of the agricultural machinery vehicle.

[0069] Obstacle categories include fixed obstacles or moving people. When the obstacle category is a fixed obstacle, the path is replanned according to the location of the obstacle. When the obstacle category is a moving person, you can stop and wait. The decision result of the decision layer is to stop and wait. The control layer converts the decision result of the decision layer into a control instruction. At this time, the control instruction is a stop and wait instruction. After the execution layer receives the stop and wait instruction, it executes the stop and wait action. When the moving person is not in front of the agricultural machinery vehicle, it continues to drive.

[0070] After obtaining the target steering angle of the agricultural machinery steering wheel based on the improved pure tracking algorithm, it is necessary to control the electric steering wheel to steer quickly and accurately, especially when the path changes sharply or turns, to quickly make precise direction adjustments to avoid path deviation caused by response lag. Therefore, this application adopts a fuzzy PID control algorithm to achieve rapid response and precise control of the steering wheel by adaptively adjusting the control parameters, ensuring that the agricultural machinery can always run efficiently and stably along the predetermined path in high-precision agricultural tasks.

[0071] The execution layer is also used to collect the actual front wheel steering angle of the agricultural machinery and feed the actual front wheel steering angle back to the control layer; the control layer is used to: use a single chip microcomputer and a fuzzy PID control algorithm to perform steering control according to the actual front wheel steering angle and the target steering angle to achieve closed-loop control. The actual front wheel steering angle is measured by an angle sensor.

[0072] like Figure 5 As shown, turn the target to the angle value X target The actual front wheel steering angle X out The steering angle error e and its error change rate de / dt are used as the input of the fuzzy controller. The fuzzy controller first fuzzifies the input, then makes inferences based on fuzzy rules, and finally defuzzifies the result of fuzzy inference to output the three parameters of the PID controller. Kp , Ki , Kd , thereby achieving adaptive tuning of PID controller parameters.

[0073] This application sets the fuzzy domain of the steering angle deviation and its rate of change to [-90°, 90°], the quantization factor is 1 / 90, and the steering angle deviation is divided into seven fuzzy language variables: big negative (NB), medium negative (NM), small negative (NS), zero (ZO), small positive (PS), medium positive (PM) and big positive (PB). This application uses a triangular membership function to represent the membership of the fuzzy set. When the system deviation is large, in order to improve the system response speed, a larger value needs to be set. Kp and Kd value, and usually Ki Set to 0 to avoid excessive integration effect. When the deviation is medium, it needs to be appropriately reduced. Kp value to reduce system overshoot and reduce Kd value to control the response speed of the system, Ki Take a moderate value to balance response speed and stability. When the deviation is small, reduce Kp and Ki value to improve the stability and anti-interference ability of the system. At the same time, in order to avoid oscillation after the system reaches the target value, it is necessary to dynamically adjust the deviation according to the size of the deviation change rate. Kd . When the rate of change is large, it decreases Kd On the contrary, increase Kd Value. Based on the above control strategy, this application has established Kp , Ki , Kd The fuzzy control rule table is shown in Table 2-4.

[0074] Table 2 KpFuzzy Control Rules

[0075] Table 3 Ki Fuzzy Control Rules

[0076] Table 4 Kd Fuzzy Control Rules

[0077] The result obtained by fuzzy reasoning needs to be defuzzified to convert the fuzzy set into specific values ​​to adjust the parameters of the PID controller. The defuzzification method used in this application is the centroid method, and its calculation formula is shown below.

[0078] (12).

[0079] In the formula, is the specific value of the PID parameter, Indicates The membership value of a fuzzy set is Indicates The central value of the fuzzy control quantity of a fuzzy set, is the number of fuzzy rules. The output of the fuzzy controller Kp , Ki , Kd To adjust the parameters of the PID controller, the control formula of the PID controller is shown below.

[0080] (13).

[0081] in, Output for steering angle control of agricultural machinery steering wheel. is the steering angle deviation.

[0082] The data processing and decision-making of this application are carried out on the edge computing platform, and the decision results are sent to the single-chip microcomputer to control the steering and braking of the agricultural machinery. The system can give full play to the advantages of the edge computing platform in real-time processing of complex algorithms and large amounts of data and the performance of the single-chip microcomputer in efficient control. The combination of the two can realize a more flexible and intelligent agricultural machinery navigation function, and improve the applicability and scalability of the system.

[0083] Regarding the improved pure tracking algorithm, the polynomial relationship between speed, path curvature and tracking error is used to dynamically calculate the forward distance in the pure tracking model, and the polynomial coefficients are optimized through the ant colony algorithm, so that the system can flexibly adapt to different vehicle speeds, path curvatures and tracking error changes. The improved pure tracking model can quickly and accurately calculate the front wheel steering angle, can achieve accurate path tracking, and can flexibly adapt to different vehicle speeds, path curvatures and tracking error changes.

[0084] The fuzzy PID control algorithm is used to control the rotation of the electric steering wheel to achieve fast and effective adjustment of the front wheel steering angle. Fuzzy PID control combines the advantages of fuzzy logic and traditional PID control, and can adaptively adjust control parameters in complex and changing environments, significantly improving the response speed and accuracy of the system.

[0085] The agricultural machinery path tracking control method based on improved pure pursuit and fuzzy PID double closed loop proposed in this application has excellent adaptability and high stability to complex field operation environments. This method can efficiently and accurately complete the full-plot automatic navigation task under a variety of operating conditions, meet the high-precision path tracking requirements of agricultural machinery in actual operations, and improve the operating efficiency and reliability of the system.

[0086] Based on the same inventive concept, the embodiment of the present application also provides an agricultural machinery intelligent navigation method based on the agricultural machinery intelligent navigation system involved above. The implementation scheme for solving the problem provided by the method is similar to the implementation scheme recorded in the above system, so the specific limitations in one or more agricultural machinery intelligent navigation method embodiments provided below can refer to the limitations on the agricultural machinery intelligent navigation system above, and will not be repeated here.

[0087] In an exemplary embodiment, Figure 6 As shown, an agricultural machinery intelligent navigation method is provided, including the following steps S1 to S4.

[0088] Step S1: Acquire environmental perception data; the environmental perception data includes agricultural machinery position information, vehicle front image, three-dimensional point cloud information and vehicle posture information.

[0089] Step S2: using an edge computing platform to determine obstacle information according to the image in front of the vehicle and the three-dimensional point cloud information; the obstacle information includes obstacle category and obstacle distance; performing path planning according to the obstacle information and the agricultural machinery position information to obtain a path planning result; using an improved pure tracking algorithm to perform path tracking decision control according to the path planning result, and calculating the target steering angle of the agricultural machinery steering wheel; the improved pure tracking algorithm uses vehicle speed, path curvature and tracking error to calculate the pre-anchor point distance.

[0090] Step S3: using a single-chip microcomputer to convert the decision result of the decision layer into a control instruction; the decision result includes the path planning result and the target steering angle of the steering wheel of the agricultural machinery.

[0091] Step S4: receiving control instructions from the control layer, and controlling the agricultural machinery to execute actions corresponding to the control instructions.

[0092] The agricultural machinery intelligent navigation method further comprises: Collecting the actual front wheel steering angle of the agricultural machinery and feeding back the actual front wheel steering angle to the control layer; By using a single chip microcomputer and a fuzzy PID control algorithm, steering control is performed according to the actual front wheel steering angle and the target steering angle to achieve closed-loop control.

[0093] As an optional implementation, the calculation formula of the pre-anchor point distance is shown in the following formula.

[0094] (14).

[0095] in, represents the pre-anchor distance, is the speed of the agricultural machinery, is the path tracking error, is the path curvature, are the vehicle speed coefficient, path tracking error coefficient and path curvature coefficient respectively, is a constant term.

[0096] As an optional implementation, path tracking decision control is performed according to the path planning result to calculate the target steering angle of the agricultural machine steering wheel, specifically including: The polynomial coefficients are optimized by using an ant colony algorithm to obtain optimal polynomial coefficients; the polynomial coefficients include a vehicle speed coefficient, a path tracking error coefficient, and a path curvature coefficient; Calculating the pre-anchor point distance according to the optimal polynomial coefficients; According to the pre-anchor point distance, a target steering angle of the agricultural machinery steering wheel is calculated.

[0097] Among them, the calculation formula for the target steering angle of the agricultural machinery steering wheel is shown in the following formula.

[0098] (15).

[0099] In the formula, for The target steering angle of the agricultural machinery steering wheel at any moment, is the vehicle wheelbase, for The angle between the pre-anchor point and the vehicle body at all times.

[0100] The technical features of the above embodiments may be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0101] This article uses specific examples to illustrate the principles and implementation methods of this application. The description of the above embodiments is only used to help understand the method and core ideas of this application. At the same time, for those skilled in the art, according to the ideas of this application, there will be changes in the specific implementation methods and application scope. In summary, the content of this specification should not be understood as limiting this application.

Claims

1. An intelligent navigation system for agricultural machinery, characterized in that: The agricultural machinery intelligent navigation system comprises: The perception layer is used to obtain environmental perception data; the environmental perception data includes agricultural machinery position information, vehicle front image, three-dimensional point cloud information and vehicle posture information; The decision layer is used to: use an edge computing platform to determine obstacle information according to the image in front of the vehicle and the three-dimensional point cloud information; the obstacle information includes obstacle category and obstacle distance; perform path planning according to the obstacle information and the agricultural machinery position information to obtain a path planning result; use an improved pure tracking algorithm to perform path tracking decision control according to the path planning result, and calculate the target steering angle of the steering wheel of the agricultural machinery; the improved pure tracking algorithm uses vehicle speed, path curvature and tracking error to calculate the pre-anchor point distance; The control layer is used to: use a single-chip microcomputer to convert the decision results of the decision layer into control instructions; the decision results include path planning results and the target steering angle of the agricultural machinery steering wheel; The execution layer is used to receive control instructions from the control layer and control the agricultural machinery to execute actions corresponding to the control instructions.

2. The intelligent navigation system for agricultural machinery according to claim 1, characterized in that: The execution layer is also used to collect the actual front wheel steering angle of the agricultural machine and feed the actual front wheel steering angle back to the control layer; The control layer is used to: utilize a single chip microcomputer and a fuzzy PID control algorithm to perform steering control according to the actual front wheel steering angle and the target steering angle to achieve closed-loop control.

3. The intelligent navigation system for agricultural machinery according to claim 1, characterized in that: The calculation formula of the pre-anchor point distance is as follows: ； in, represents the pre-anchor distance, is the speed of the agricultural machinery, is the path tracking error, is the path curvature, are the vehicle speed coefficient, path tracking error coefficient and path curvature coefficient respectively, is a constant term.

4. The intelligent navigation system for agricultural machinery according to claim 3, characterized in that: In adopting the improved pure tracking algorithm, performing path tracking decision control according to the path planning result, and calculating the target steering angle of the agricultural machinery steering wheel, the decision layer is used to: The polynomial coefficients are optimized by using an ant colony algorithm to obtain optimal polynomial coefficients; the polynomial coefficients include a vehicle speed coefficient, a path tracking error coefficient, and a path curvature coefficient; Calculating the pre-anchor point distance according to the optimal polynomial coefficients; According to the pre-anchor point distance, a target steering angle of the agricultural machinery steering wheel is calculated.

5. The intelligent navigation system for agricultural machinery according to claim 4, characterized in that: The calculation formula of the target steering angle of the agricultural machinery steering wheel is as follows: ； In the formula, for The target steering angle of the agricultural machinery steering wheel at any moment, is the vehicle wheelbase, for The angle between the pre-anchor point and the vehicle body at all times.

6. An agricultural machinery intelligent navigation method based on the agricultural machinery intelligent navigation system according to any one of claims 1 to 5, characterized in that: The agricultural machinery intelligent navigation method comprises: Acquire environmental perception data; the environmental perception data includes agricultural machinery position information, vehicle front image, three-dimensional point cloud information and vehicle posture information; An edge computing platform is used to determine obstacle information according to the image in front of the vehicle and the three-dimensional point cloud information; the obstacle information includes obstacle category and obstacle distance; path planning is performed according to the obstacle information and the agricultural machinery position information to obtain a path planning result; an improved pure tracking algorithm is used to perform path tracking decision control according to the path planning result, and a target steering angle of the steering wheel of the agricultural machinery is calculated; the improved pure tracking algorithm uses vehicle speed, path curvature and tracking error to calculate the pre-anchor point distance; The single chip microcomputer is used to convert the decision results of the decision layer into control instructions; the decision results include the path planning results and the target steering angle of the agricultural machinery steering wheel; Receive control instructions from the control layer and control the agricultural machinery to execute actions corresponding to the control instructions.

7. The agricultural machinery intelligent navigation method according to claim 6, characterized in that: The agricultural machinery intelligent navigation method further comprises: Collecting the actual front wheel steering angle of the agricultural machinery and feeding back the actual front wheel steering angle to the control layer; By using a single chip microcomputer and a fuzzy PID control algorithm, steering control is performed according to the actual front wheel steering angle and the target steering angle to achieve closed-loop control.

8. The agricultural machinery intelligent navigation method according to claim 6, characterized in that: The calculation formula of the pre-anchor point distance is as follows: ； in, represents the pre-anchor distance, is the speed of the agricultural machinery, is the path tracking error, is the path curvature, are the vehicle speed coefficient, path tracking error coefficient and path curvature coefficient respectively, is a constant term.

9. The agricultural machinery intelligent navigation method according to claim 8, characterized in that: The improved pure tracking algorithm is used to perform path tracking decision control according to the path planning result, and the target steering angle of the agricultural machinery steering wheel is calculated, specifically including: The polynomial coefficients are optimized by using an ant colony algorithm to obtain optimal polynomial coefficients; the polynomial coefficients include a vehicle speed coefficient, a path tracking error coefficient, and a path curvature coefficient; Calculating the pre-anchor point distance according to the optimal polynomial coefficients; According to the pre-anchor point distance, a target steering angle of the agricultural machinery steering wheel is calculated.

10. The agricultural machinery intelligent navigation method according to claim 9, characterized in that: The calculation formula of the target steering angle of the agricultural machinery steering wheel is as follows: ； In the formula, for The target steering angle of the agricultural machinery steering wheel at any moment, is the vehicle wheelbase, for The angle between the pre-anchor point and the vehicle body at all times.

Images