Agricultural Machinery Autopilot Control Method and System Based on Intelligent Cockpit

Abstract

The present invention discloses an agricultural machinery automatic driving control method and system based on an intelligent cockpit. The method includes: generating a grid map of a working area, global path planning, updating the grid map, local path planning, and automatic driving control. The present invention belongs to the technical field of automatic driving, specifically referring to an agricultural machinery automatic driving control method and system based on an intelligent cockpit. This solution dynamically generates particle positions by combining trigonometric functions and random numbers, calculates fitness values according to an obstacle avoidance penalty factor and a path length penalty factor, updates particle positions according to a deviation coefficient, and determines an optimal global path; a value set is formulated based on four constraints, and then candidate speed combinations are obtained by resampling. Local paths under different candidate speed combinations are simulated, and the local paths are comprehensively evaluated from five aspects, and the optimal local path with the maximum comprehensive evaluation value is selected to determine an optimal path planning scheme, ensuring the stability and reliability of the automatic driving of agricultural machinery.

Description

Technical Field

[0001] The present invention belongs to the technical field of autonomous driving, and specifically refers to an agricultural machinery autonomous driving control method and system based on an intelligent cockpit. Background Art

[0002] The agricultural machinery autonomous driving control method uses advanced sensor technology, machine learning algorithms, and data analysis to perceive the working area environment in real time, and controls the agricultural machinery to perform autonomous driving through the intelligent cockpit to improve the operation efficiency and accuracy. However, in the existing agricultural machinery autonomous driving control methods, there are problems such as the irregular shape and complex terrain of the agricultural machinery working area, and the existing path planning algorithms cannot fully and effectively utilize the working area for global path search, and cannot correctly select the optimal global path, resulting in the inability of the agricultural machinery to efficiently and safely complete the operation, reducing the working efficiency and increasing the operation cost; in the existing agricultural machinery autonomous driving control methods, when the agricultural machinery encounters moving obstacles in the working area, it cannot effectively find the optimal local path, and the agricultural machinery collides with the obstacles and deviates too much from the global path due to avoiding the obstacles, resulting in damage to the agricultural machinery and increasing energy consumption, reducing power efficiency, and increasing operation cost. Summary of the Invention

[0003] In view of the above situation, to overcome the defects of the prior art, the present invention provides an agricultural machinery automatic driving control method and system based on an intelligent cockpit. In the existing agricultural machinery automatic driving control methods, the shape of the working area of agricultural machinery is irregular and the terrain is complex. The existing path planning algorithms cannot make full and effective use of the working area for global path search, and cannot correctly select the optimal global path, resulting in the inability of agricultural machinery to complete operations efficiently and safely, reducing work efficiency and increasing operation costs. In this solution, a grid map is used as the search space, and trigonometric functions and random numbers are combined to dynamically generate particle positions, enhancing the scope and breadth of global path search; the fitness value is calculated according to the obstacle avoidance penalty factor and the path length penalty factor to reduce the collision risk and improve the operation efficiency; the particle position is updated according to the deviation coefficient to determine the optimal global path, enhancing the diversity and flexibility of particles in the search space and increasing the chance of finding the optimal global path, ensuring the stability and reliability of agricultural machinery automatic driving; in the existing agricultural machinery automatic driving control methods, when agricultural machinery encounters moving obstacles in the working area, it cannot effectively find the optimal local path, and the agricultural machinery collides with the obstacles and deviates too much from the global path due to avoiding the obstacles, resulting in damage to the agricultural machinery, increased energy consumption, reduced power efficiency, and increased operation costs. In this solution, a value set is formulated based on the value ranges of linear velocity, angular velocity, minimum turning radius, acceleration and deceleration limits, and braking distance limits, and candidate speed combinations are obtained by resampling, avoiding the loss of control of agricultural machinery or the inability to drive along the planned path due to unreasonable speeds; the local paths under different candidate speed combinations are simulated, and the local paths are comprehensively evaluated from five aspects: safety, power efficiency, target proximity, path consistency, and direction consistency, and the optimal local path with the maximum comprehensive evaluation value is selected, enabling the agricultural machinery to avoid moving obstacles in the optimal way and minimizing the adverse effects on the operation efficiency, safety, and overall driving trajectory of the agricultural machinery.

[0004] The technical solution adopted by the present invention is as follows: The agricultural machinery automatic driving control method based on an intelligent cockpit provided by the present invention includes the following steps:

[0005] Step S1: Generate a grid map of the working area;

[0006] Step S2: Global path planning;

[0007] Step S3: Update the grid map;

[0008] Step S4: Local path planning;

[0009] Step S5: Automatic driving control.

[0010] Further, in step S1, the generation of the grid map of the working area specifically includes the following steps:

[0011] Step S11: Working area image acquisition; acquire the working area image through the sensor installed on the agricultural machine, and transmit the acquired working area image to the intelligent cockpit;

[0012] Step S12: Image preprocessing; in the intelligent cockpit, preprocess the acquired working area image, use the histogram equalization algorithm to enhance the contrast of the working area image, and adopt the Gaussian filtering algorithm to remove the noise in the working area image;

[0013] Step S13: Image recognition; in the intelligent cockpit, based on the Canny edge detection algorithm, identify the boundary data of the working area from the working area image, and use the YOLO model to identify the obstacle data in the working area image;

[0014] Step S14: Construct a grid map; according to the boundary data and obstacle data identified from the working area image, construct a grid map of the working area in the intelligent cockpit, and calibrate the starting point and the target point.

[0015] Further, in step S2, the global path planning specifically includes the following steps:

[0016] Step S21: Initial global path; in the intelligent cockpit, use the grid map of the working area as the search space for the global path, initialize M particle positions between the starting point and the target point, and each particle position represents a different candidate global path; the initial particle positions are generated by combining the periodic perturbation of the trigonometric function and random numbers, and introducing an adjustment factor to dynamically adjust the mapping method in different intervals to generate M mapping values, and then determine M initial particle positions in the search space. The formula used is as follows:

[0017] ;

[0018] ;

[0019] In the formula, is the m-th initial particle position, q m+1 and q m are the mapping values corresponding to the (m + 1)-th and m-th initial particle positions respectively, S lb and S ub are the lower and upper limits of the search space respectively, α is an adjustment factor within the range of (0, 1), mod(·), sin(·) and cos(·) are the modulo function, sine function and cosine function respectively, r 1 and r 2 are random numbers;

[0020] Step S22: Calculate the fitness value, and preset the safety distance , calculate the obstacle avoidance penalty factor and the path length penalty factor of the candidate global path corresponding to each particle position, and then perform weighted processing according to the weight coefficient to obtain the fitness value of each particle position;

[0021] Step S23: Design the deviation coefficient, and the formula used is as follows:

[0022] ;

[0023] In the formula, is the deviation coefficient of the m-th particle position at the u-th search, and are the minimum fitness value and the average fitness value at the u-th search respectively, β is the perturbation amplitude, u max is the maximum number of searches, is the fitness value of the m-th particle position at the u-th search;

[0024] Step S24: Update the particle position; update the particle position according to the deviation coefficient, and the formula used is as follows:

[0025] ;

[0026] In the formula, and are the m-th particle positions at the (u + 1)-th and u-th searches respectively, r 3 、r 4 and r 5 are random numbers, is a position randomly selected from the M particle positions at the u-th search, δ is the attraction coefficient, is the global optimal position at the u-th search, and the global optimal position is the particle position with the minimum fitness value, k 1 and k 2 are the golden section coefficients;

[0027] Step S25: Determine the optimal global path; preset the fitness threshold, update the fitness value of the particle position. If the fitness value of the global optimal position is lower than the fitness threshold, then use the candidate global path represented by the global optimal position as the optimal global path, and the intelligent cockpit controls the agricultural machinery to start from the starting point and perform autonomous driving according to the optimal global path; otherwise, if the maximum number of searches is reached, return to Step S21 to re-initialize the global path; otherwise, return to Step S23 to continue the search.

[0028] Further, in step S3, when updating the grid map, during the process of the agricultural machine driving autonomously along the optimal global path, an observation area V is delimited with the agricultural machine as the center and 3 times the safety distance as the radius. The agricultural machine monitors the environment within the observation area V in real time through sensors and updates the grid map. If a moving obstacle is detected within the observation area V, local path planning is performed; otherwise, the agricultural machine will continue to drive autonomously along the optimal global path.

[0029] Further, in step S4, the local path planning specifically includes the following steps:

[0030] Step S41: Generate candidate speed combinations, including the following steps:

[0031] Step S411: Speed limit; Based on the minimum and maximum values of the linear velocity v and angular velocity ω of the agricultural machine, a first value set of speed combinations is formulated ; where, v min and v max are respectively the minimum and maximum values of the linear velocity, ω min and ω max are respectively the minimum and maximum values of the angular velocity;

[0032] Step S412: Minimum turning radius limit; Based on the minimum turning radius of the agricultural machine, a second value set of speed combinations is formulated ; where, is the minimum turning radius;

[0033] Step S413: Acceleration and deceleration limit; Since the torque of the drive motor is limited, the agricultural machine is restricted during acceleration and deceleration, and a third value set of speed combinations is formulated ; where, v t and ω t are respectively the linear velocity and angular velocity of the agricultural machine at the current moment t, and are respectively the maximum deceleration value and maximum acceleration value of the linear velocity, and are respectively the maximum deceleration value and maximum acceleration value of the angular velocity, is the time interval;

[0034] Step S414: Braking distance limit; In the case of maximum deceleration, a fourth value set of speed combinations at which the agricultural machine can stop before colliding with the obstacle is formulated ; where, is the distance between the trajectory traveled by the agricultural machine at the speed combination (v, ω) and the nearest obstacle;

[0035] Step S415: Sampling; Obtain the final value set of the agricultural machine speed combination , for the linear velocity and angular velocity in and step sizes respectively, sample to obtain candidate linear velocity values and candidate angular velocity values, pair the candidate linear velocity values and candidate angular velocity values to obtain N v ×N ω candidate velocity combinations (v, ω);

[0036] Step S42: Simulate the local path; when the agricultural machine detects a moving obstacle in the observation area V, collect the current position (x, y) and direction angle θ of the agricultural machine, and simulate the local path of the agricultural machine according to each candidate velocity combination (v, ω) to obtain the positions v ×N ω of the ends of N local paths J(v, ω) and the direction angle ;

[0037] Step S43: Local path evaluation, including the following steps:

[0038] Step S431: Safety evaluation; obtain the safety value of J(v, ω) through the distances between the ends of the local path J(v, ω) and each obstacle in the observation area V ; the formula used is as follows:

[0039] ;

[0040] ;

[0041] In the formula, E is the number of obstacles detected in the observation area V, γ e is the safety factor of the e-th obstacle, is the distance between the end of the local path J(v, ω) and the e-th obstacle, and σ max is the maximum distance error value;

[0042] Step S432: Power benefit evaluation; obtain the power benefit value of J(v, ω) by considering the acceleration at the end of the local path J(v, ω) and the maximum acceleration ability of the agricultural machine ; the formula used is as follows:

[0043] ;

[0044] In the formula, and are the linear acceleration and angular acceleration at the end of the local path J(v, ω) respectively;

[0045] Step S433: Target proximity evaluation; by considering the distance between the end of the local path J(v, ω) and the target point , the target proximity value of J(v, ω) is obtained ;

[0046] Step S434: Path consistency evaluation; by considering the distance between the end of the local path J(v, ω) and the nearest path point in the global path , the path consistency value of J(v, ω) is obtained ;

[0047] Step S435: Direction consistency evaluation; by considering the angle difference between the direction angle at the end of the local path J(v, ω) and the direction angle from the agricultural machine to the target point , the direction consistency value of J(v, ω) is obtained ;

[0048] Step S436: Total evaluation; normalize the safety value, power benefit value, target proximity value, path consistency value and direction consistency value respectively, and then perform weighted combination to obtain the total evaluation value of the local path ;

[0049] Step S44: Determine the optimal local path; select the local path with the maximum comprehensive evaluation value from all the simulated local paths as the optimal local path, and the intelligent cockpit controls the agricultural machine to perform automatic driving along the optimal local path to avoid moving obstacles.

[0050] Furthermore, in step S5, the automatic driving control is that after the intelligent cockpit controls the agricultural machine to reach the end of the optimal local path, the grid map is updated in real time. If a moving obstacle is detected again in the observation area V, local path planning is performed again; otherwise, the agricultural machine continues to perform automatic driving along the optimal global path until the agricultural machine reaches the target point and the automatic driving ends.

[0051] The agricultural machine automatic driving control system based on the intelligent cockpit provided by the present invention includes a grid map module for generating a working area, a global path planning module, a grid map updating module, a local path planning module and an automatic driving control module;

[0052] The grid map module for generating the working area collects the working area image and performs preprocessing, edge detection and obstacle recognition in the intelligent cockpit, constructs the grid map and calibrates the starting point and the target point, and sends the data to the global path planning module;

[0053] The global path planning module uses the grid map as the search space, dynamically generates particle positions by combining trigonometric functions and random numbers, calculates the fitness value according to the obstacle avoidance penalty factor and the path length penalty factor, updates the particle positions according to the deviation coefficient, determines the optimal global path, and the intelligent cockpit controls the agricultural machinery to drive automatically according to the optimal global path, and sends the data to the updated grid map module;

[0054] The updated grid map module is for the agricultural machinery to monitor the observation area in real time and update the grid map, conduct local path planning when detecting moving obstacles, otherwise continue to drive along the optimal global path, and send the data to the local path planning module;

[0055] The local path planning module formulates a value set based on the value ranges of linear velocity, angular velocity, minimum turning radius, acceleration and deceleration limits, and braking distance limits, resamples to obtain candidate speed combinations, simulates local paths under different candidate speed combinations, and comprehensively evaluates the local paths from five aspects: safety, power efficiency, target proximity, path consistency, and direction consistency, selects the optimal local path with the maximum comprehensive evaluation value, and sends the data to the automatic driving control module;

[0056] The automatic driving control module is for the intelligent cockpit to control the agricultural machinery to drive along the optimal local path, and update the grid map in real time. If a moving obstacle is detected, it will re-plan the local path, otherwise continue to drive along the optimal global path until it reaches the target point and ends.

[0057] The beneficial effects achieved by the present invention using the above solution are as follows:

[0058] (1) Aiming at the problems in the existing agricultural machinery automatic driving control methods that the shape of the working area of the agricultural machinery is irregular and the terrain is complex, the existing path planning algorithms cannot fully and effectively utilize the working area for global path search, and cannot correctly select the optimal global path, resulting in the agricultural machinery being unable to complete operations efficiently and safely, reducing work efficiency and increasing operation costs. This solution uses the grid map as the search space, dynamically generates particle positions by combining trigonometric functions and random numbers, increases the diversity of candidate global paths, and enhances the scope and breadth of global path search; calculates the fitness value according to the obstacle avoidance penalty factor and the path length penalty factor to ensure the safety and efficiency of the selected path, reduce the collision risk, and improve the operation efficiency; updates the particle positions according to the deviation coefficient, determines the optimal global path, enhances the diversity and flexibility of the particles in the search space, increases the chance of finding the optimal global path, ensures the stability and reliability of the agricultural machinery automatic driving, reduces the safety risk caused by global path planning mistakes, and improves the operation quality.

[0059] (2)In view of the problems existing in the existing agricultural machinery automatic driving control method, such as when the agricultural machinery encounters moving obstacles in the working area, it cannot effectively find the optimal local path, the agricultural machinery collides with the obstacles, and deviates too much from the global path due to avoiding the obstacles, resulting in damage to the agricultural machinery, increased energy consumption, reduced power efficiency, and increased operation costs. This solution formulates a value set based on the value ranges of linear velocity, angular velocity, minimum turning radius, acceleration and deceleration limits, and braking distance limits, and then samples to obtain candidate speed combinations to avoid the agricultural machinery getting out of control or being unable to travel along the planned path due to unreasonable speeds. Simulate the local paths under different candidate speed combinations, comprehensively evaluate the local paths from five aspects: safety, power efficiency, target proximity, path consistency, and direction consistency, and select the optimal local path with the maximum comprehensive evaluation value, so that the agricultural machinery can avoid moving obstacles in the optimal way and minimize the adverse effects on the operation efficiency, safety, and overall driving trajectory of the agricultural machinery. Description of the Drawings

[0060] Figure 1 It is a schematic flow chart of the agricultural machinery automatic driving control method based on an intelligent cockpit provided by the present invention;

[0061] Figure 2 It is a schematic diagram of the agricultural machinery automatic driving control system based on an intelligent cockpit provided by the present invention;

[0062] Figure 3 It is a schematic flow chart of step S2;

[0063] Figure 4 It is a schematic flow chart of step S4.

[0064] The drawings are used to provide a further understanding of the present invention, and constitute a part of the specification. They are used together with the embodiments of the present invention to explain the present invention, and do not constitute a limitation to the present invention. Detailed Embodiments

[0065] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments; based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0066] In the description of the present invention, it should be understood that the terms "upper", "lower", "front", "rear", "left", "right", "top", "bottom", "inner", "outer", etc. indicating the orientation or positional relationship are based on the orientation or positional relationship shown in the drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore should not be construed as a limitation to the present invention.

[0067] Embodiment 1. Refer to Figure 1 , the agricultural machinery automatic driving control method based on an intelligent cockpit provided by the present invention includes the following steps:

[0068] Step S1: Generate a grid map of the working area; collect images of the working area and perform preprocessing, edge detection, and obstacle recognition in the intelligent cockpit, construct a grid map, and calibrate the starting point and the target point;

[0069] Step S2: Global path planning; use the grid map as the search space, dynamically generate particle positions by combining trigonometric functions and random numbers, calculate the fitness value according to the obstacle avoidance penalty factor and the path length penalty factor, update the particle positions according to the deviation coefficient, determine the optimal global path, and the intelligent cockpit controls the agricultural machinery to perform automatic driving according to the optimal global path;

[0070] Step S3: Update the grid map; the agricultural machinery monitors the observation area in real time and updates the grid map, and performs local path planning when a moving obstacle is detected, otherwise continues to drive along the optimal global path;

[0071] Step S4: Local path planning; formulate a value set based on the value ranges of the linear velocity, angular velocity, minimum turning radius, acceleration and deceleration limits, and braking distance limit, then resample to obtain candidate speed combinations, simulate the local paths under different candidate speed combinations, and comprehensively evaluate the local paths from five aspects: safety, power efficiency, target proximity, path consistency, and direction consistency, and select the optimal local path with the maximum comprehensive evaluation value;

[0072] Step S5: Automatic driving control; the intelligent cockpit controls the agricultural machinery to drive along the optimal local path, and updates the grid map in real time. If a moving obstacle is detected, the local path is re-planned, otherwise continues to drive along the optimal global path until the target point is reached and the process ends.

[0073] Embodiment 2. Refer to Figure 1 , based on the above embodiment, in step S1, generating the grid map of the working area specifically includes the following steps:

[0074] Step S11: Acquisition of working area image; The working area image is acquired by a sensor installed on the agricultural machine, and the acquired working area image is transmitted to the intelligent cockpit;

[0075] Step S12: Image preprocessing; Inside the intelligent cockpit, the acquired working area image is preprocessed. The histogram equalization algorithm is used to enhance the contrast of the working area image, and the Gaussian filtering algorithm is adopted to remove the noise in the working area image;

[0076] Step S13: Image recognition; Inside the intelligent cockpit, based on the Canny edge detection algorithm, the boundary data of the working area are recognized from the working area image, and the YOLO model is used to recognize the obstacle data in the working area image;

[0077] Step S14: Construction of grid map; According to the boundary data and obstacle data recognized from the working area image, a grid map of the working area is constructed inside the intelligent cockpit, and the starting point and the target point are calibrated.

[0078] Embodiment 3, refer to Figure 1 and Figure 3 , based on the above embodiment, in step S2, the global path planning specifically includes the following steps:

[0079] Step S21: Initial global path; Inside the intelligent cockpit, the grid map of the working area is used as the search space for the global path. M particle positions are initialized between the starting point and the target point, and each particle position represents a different candidate global path; The initial particle positions are generated by combining the periodic perturbation of the trigonometric function and random numbers, and introducing an adjustment factor to dynamically adjust the mapping method in different intervals to generate M mapping values, and then M initial particle positions are determined in the search space. The formula used is as follows:

[0080] ;

[0081] ;

[0082] In the formula, is the m-th initial particle position, q m+1 and q m are the mapping values corresponding to the (m + 1)-th and m-th initial particle positions respectively, S lb and S ub are the lower and upper limits of the search space respectively, α is an adjustment factor within the range of (0, 1), mod(·), sin(·) and cos(·) are the modulo function, sine function and cosine function respectively, r 1 and r 2 are random numbers within the range of (0, 1);

[0083] Step S22: Calculate the fitness value and preset the safety distance , calculate the obstacle avoidance penalty factor and the path length penalty factor of the candidate global path corresponding to each particle position, and then perform weighted processing according to the weight coefficients to obtain the fitness value of each particle position. The formula used is as follows:

[0084] ;

[0085] ;

[0086] ;

[0087] In the formula, is the fitness value of the m-th particle position at the u-th search, and are the obstacle avoidance penalty factor and the path length penalty factor of the m-th particle position at the u-th search respectively, and are the weight coefficients of the obstacle avoidance penalty factor and the path length penalty factor respectively, and , is the distance between the m-th particle position and the nearest static obstacle at the u-th search, is the length of the global path represented by the m-th particle position at the u-th search, is the maximum length among the global paths represented by the M particle positions at the u-th search;

[0088] Step S23: Design the deviation coefficient. The formula used is as follows:

[0089] ;

[0090] In the formula, is the deviation coefficient of the m-th particle position at the u-th search, and are the minimum fitness value and the average fitness value at the u-th search respectively, β is the perturbation amplitude, and u max is the maximum number of searches;

[0091] Step S24: Update the particle position; update the particle position according to the deviation coefficient. The formula used is as follows:

[0092] ;

[0093] In the formula, and are the m-th particle positions at the (u + 1)-th and u-th searches respectively, and r 3 、r 4 and r 5 are random numbers within the range of (0, 1), is a position randomly selected from the positions of M particles at the u-th search. δ is the attraction coefficient, which is used to control the attraction intensity to is the global optimal position at the u-th search. The global optimal position is the particle position with the minimum fitness value. k 1 and k 2 are the golden section coefficients;

[0094] Step S25: Determine the optimal global path; preset a fitness threshold, update the fitness value of the particle position. If the fitness value of the global optimal position is lower than the fitness threshold, then use the candidate global path represented by the global optimal position as the optimal global path, and the intelligent cockpit controls the agricultural machinery to start from the starting point and perform autonomous driving according to the optimal global path; otherwise, if the maximum search number is reached, return to Step S21 to re-initialize the global path; otherwise, return to Step S23 to continue the search.

[0095] By performing the above operations, aiming at the problems in the existing agricultural machinery autonomous driving control method that the shape of the working area of the agricultural machinery is irregular and the terrain is complex, the existing path planning algorithms cannot make full and effective use of the working area for global path search, cannot correctly select the optimal global path, resulting in the agricultural machinery being unable to complete the operation efficiently and safely, reducing the work efficiency and increasing the operation cost. In this solution, the grid map is used as the search space, combined with trigonometric functions and random numbers to dynamically generate particle positions, increasing the diversity of candidate global paths, enhancing the scope and breadth of global path search; calculating the fitness value according to the obstacle avoidance penalty factor and the path length penalty factor to ensure the safety and efficiency of the selected path, reducing the collision risk and improving the operation efficiency; updating the particle positions according to the deviation coefficient, determining the optimal global path, enhancing the diversity and flexibility of the particles in the search space, increasing the chance of finding the optimal global path, ensuring the stability and reliability of the agricultural machinery autonomous driving, reducing the safety risk caused by the global path planning error, and improving the operation quality.

[0096] Embodiment 4, refer to Figure 1 , based on the above embodiment, in Step S3, updating the grid map is during the process of the agricultural machinery performing autonomous driving along the optimal global path. Taking the agricultural machinery as the center and 3 times the safety distance as the radius, delimit an observation area V. The agricultural machinery monitors the environment in the observation area V in real time through sensors and updates the grid map; if a moving obstacle is detected in the observation area V, then perform local path planning; otherwise, the agricultural machinery will continue to perform autonomous driving along the optimal global path.

[0097] Embodiment 5, refer to Figure 1 and Figure 4, based on the above embodiment, in step S4, the local path planning specifically includes the following steps:

[0098] Step S41: Generate candidate speed combinations, including the following steps:

[0099] Step S411: Speed limit; Based on the minimum and maximum values of the linear velocity v and angular velocity ω of the agricultural machine, formulate the first value set of the speed combination ; where, v min and v max are respectively the minimum and maximum values of the linear velocity, ω min and ω max are respectively the minimum and maximum values of the angular velocity;

[0100] Step S412: Minimum turning radius limit; Based on the minimum turning radius of the agricultural machine, formulate the second value set of the speed combination ; where, is the minimum turning radius;

[0101] Step S413: Acceleration and deceleration limit; Since the torque of the drive motor is limited, the agricultural machine is restricted during acceleration and deceleration, and the third value set of the speed combination is formulated ; where, v t and ω t are respectively the linear velocity and angular velocity of the agricultural machine at the current moment t, and are respectively the maximum deceleration value and maximum acceleration value of the linear velocity, and are respectively the maximum deceleration value and maximum acceleration value of the angular velocity, is the time interval;

[0102] Step S414: Braking distance limit; In the case of maximum deceleration, formulate the fourth value set of the speed combination at which the agricultural machine can stop before colliding with an obstacle ; where, is the distance between the trajectory traveled by the agricultural machine at the speed combination (v, ω) and the nearest obstacle;

[0103] Step S415: Sampling; Obtain the final value set of the agricultural machine speed combination , for in the linear velocity and angular velocity are sampled respectively with step sizes and to obtain candidate linear velocity values and candidate angular velocity values, and pair the candidate linear velocity values and candidate angular velocity values to obtain N v ×N ω candidate speed combinations (v, ω);

[0104] Step S42: Simulate the local path; when the agricultural machine detects a moving obstacle within the observation area V, collect the current position (x, y) and the orientation angle θ of the agricultural machine, and simulate the local path of the agricultural machine according to each candidate speed combination (v, ω) to obtain the positions v ×N ω of the ends of N local paths J(v, ω) and the orientation angles ; the formulas used are as follows:

[0105] ;

[0106] ;

[0107] ;

[0108] In the formula, , and are respectively the abscissa position, ordinate position and orientation angle at the end of the simulated local path, and x, y and are respectively the abscissa position, ordinate position and orientation angle of the current agricultural machine;

[0109] Step S43: Evaluate the local path, including the following steps:

[0110] Step S431: Safety evaluation; obtain the safety value of J(v, ω) through the distances between the ends of the local paths J(v, ω) and each obstacle within the observation area V ; the formula used is as follows:

[0111] ;

[0112] ;

[0113] In the formula, E is the number of obstacles detected within the observation area V, γ e is the safety factor of the e-th obstacle, is the distance between the end of the local path J(v, ω) and the e-th obstacle, and σ max is the maximum distance error value;

[0114] Step S432: Power benefit evaluation; obtain the power benefit value of J(v, ω) by considering the acceleration at the end of the local path J(v, ω) and the maximum acceleration ability of the agricultural machine ; the formula used is as follows:

[0115] ;

[0116] In the formula, and are the linear acceleration and angular acceleration at the end of the local path J(v, ω), respectively;

[0117] Step S433: Target proximity evaluation; by considering the distance between the end of the local path J(v, ω) and the target point , the target proximity value of J(v, ω) is obtained ;

[0118] Step S434: Path consistency evaluation; by considering the distance between the end of the local path J(v, ω) and the nearest path point in the global path , the path consistency value of J(v, ω) is obtained ;

[0119] Step S435: Direction consistency evaluation; by considering the angle difference between the direction angle at the end of the local path J(v, ω) and the direction angle from the agricultural machine to the target point , the direction consistency value of J(v, ω) is obtained ;

[0120] Step S436: Total evaluation; the safety value, power benefit value, target proximity value, path consistency value, and direction consistency value are respectively normalized, and then weighted and combined to obtain the total evaluation value of the local path ; the formula used is as follows:

[0121] ;

[0122] where z 1 , z 2 , z 3 , z 4 , and z 5 are the weight coefficients of the safety value, power benefit value, target proximity value, path consistency value, and direction consistency value respectively, , , , , , and are respectively , , , , and the values after normalization;

[0123] Step S44: Determine the optimal local path; select the local path with the maximum comprehensive evaluation value from all the simulated local paths as the optimal local path, and the intelligent cockpit controls the agricultural machine to perform autonomous driving along the optimal local path to avoid moving obstacles.

[0124] By performing the above operations, in view of the problems existing in the existing agricultural machinery automatic driving control method, such as when the agricultural machinery encounters a moving obstacle in the working area, it cannot effectively find the optimal local path, the agricultural machinery collides with the obstacle, and deviates excessively from the global path due to avoiding the obstacle, resulting in damage to the agricultural machinery, increased energy consumption, reduced power efficiency, and increased operation costs. This solution formulates a value set based on the value ranges of linear velocity, angular velocity, minimum turning radius, acceleration and deceleration limits, and braking distance limits, and then samples to obtain candidate speed combinations, avoiding the loss of control of the agricultural machinery or the inability to drive along the planned path due to unreasonable speeds. Simulate the local paths under different candidate speed combinations, comprehensively evaluate the local paths from five aspects: safety, power efficiency, target proximity, path consistency, and direction consistency, and select the optimal local path with the maximum comprehensive evaluation value, enabling the agricultural machinery to avoid moving obstacles in an optimal manner and minimizing the adverse effects on the operation efficiency, safety, and overall driving trajectory of the agricultural machinery.

[0125] Embodiment Six. Refer to Figure 1 , based on the above embodiment, in step S5, the automatic driving control is that after the intelligent cockpit controls the agricultural machinery to reach the end of the optimal local path, the grid map is updated in real time. If a moving obstacle is detected again in the observation area V, local path planning is performed again; otherwise, the agricultural machinery continues to perform automatic driving along the optimal global path until the agricultural machinery reaches the target point, and the automatic driving ends.

[0126] Embodiment Seven. Refer to Figure 2 , based on the above embodiment, the agricultural machinery automatic driving control system based on the intelligent cockpit provided by the present invention includes a grid map module for generating a working area, a global path planning module, an updated grid map module, a local path planning module, and an automatic driving control module;

[0127] The grid map module for generating the working area collects the working area image and performs preprocessing, edge detection, and obstacle recognition in the intelligent cockpit, constructs the grid map, calibrates the starting point and the target point, and sends the data to the global path planning module;

[0128] The global path planning module uses the grid map as the search space, dynamically generates particle positions by combining trigonometric functions and random numbers, calculates the fitness value according to the obstacle avoidance penalty factor and the path length penalty factor, updates the particle positions according to the deviation coefficient, determines the optimal global path, and the intelligent cockpit controls the agricultural machinery to perform automatic driving according to the optimal global path, and sends the data to the updated grid map module;

[0129] The updated grid map module is that the agricultural machinery monitors the observation area in real time and updates the grid map, performs local path planning when a moving obstacle is detected, otherwise continues to drive along the optimal global path, and sends the data to the local path planning module;

[0130] The local path planning module formulates a value set based on the value ranges of the linear velocity, angular velocity, minimum turning radius, acceleration and deceleration limits, and braking distance limit, then samples to obtain candidate speed combinations, simulates local paths under different candidate speed combinations, comprehensively evaluates the local paths from five aspects: safety, power efficiency, target proximity, path consistency, and direction consistency, selects the optimal local path with the maximum comprehensive evaluation value, and sends the data to the autonomous driving control module;

[0131] The autonomous driving control module is an intelligent cockpit that controls the agricultural machinery to travel along the optimal local path, and updates the grid map in real time. If a moving obstacle is detected, it re-plans the local path; otherwise, it continues to travel along the optimal global path until it reaches the target point and ends.

[0132] It should be noted that in this article, relational terms such as first and second are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the term "comprising", "including" or any other variant thereof is intended to cover non-exclusive inclusion, so that a process, method, article or device comprising a series of elements includes not only those elements, but also other elements not expressly listed, or elements inherent to such process, method, article or device.

[0133] Although the embodiments of the present invention have been shown and described, those of ordinary skill in the art can understand that various changes, modifications, substitutions and variations can be made to these embodiments without departing from the principle and spirit of the present invention.

[0134] The above describes the present invention and its implementation manners. Such description is not restrictive. What is shown in the drawings is only one of the implementation manners of the present invention, and the actual structure is not limited thereto. Generally speaking, if those of ordinary skill in the art are inspired by it and design similar structural manners and embodiments without creative efforts without departing from the purpose of the present invention, they shall fall within the protection scope of the present invention.

Claims

1. The automatic driving control method of agricultural machinery based on the intelligent cockpit is characterized by: The method comprises the following steps: Step S1: Generate a grid map of the working area; collect images of the working area and perform preprocessing, edge detection and obstacle recognition in the smart cockpit, build a grid map and calibrate the starting point and target point; Step S2: global path planning; Step S3: updating the grid map; the agricultural machine monitors the observation area in real time and updates the grid map, and performs local path planning when a moving obstacle is detected, otherwise it continues to travel along the optimal global path; Step S4: local path planning: formulate a value set based on the range of linear speed, angular speed, minimum turning radius, acceleration and deceleration limits, and braking distance limits, and then sample to obtain candidate speed combinations, simulate local paths under different candidate speed combinations, and comprehensively evaluate the local paths from five aspects: safety, power efficiency, target proximity, path consistency, and direction consistency, and select the optimal local path with the maximum comprehensive evaluation value; Step S5: Automatic driving control; the intelligent cockpit controls the agricultural machine to drive along the optimal local path and updates the grid map in real time. If a moving obstacle is detected, the local path is replanned, otherwise it continues to drive along the optimal global path until it reaches the target point; In step S2, the global path planning specifically includes the following steps: Step S21: Initial global path: In the smart cockpit, the grid map of the working area is used as the search space of the global path, and M initial particle positions are set between the starting point and the target point, each particle position represents a different candidate global path; the initial particle position is determined by combining the periodic perturbation of the trigonometric function with random numbers, and introducing an adjustment factor to dynamically adjust the mapping method in different intervals, generating M mapping values, and then determining the M initial particle positions in the search space. The formula used is as follows: ； ； In the formula, is the initial particle position of the mth particle, q m+1 and q m are the mapping values ​​corresponding to the positions of the m+1th and mth initial particles, respectively, S lb and S ub are the lower and upper limits of the search space, α is the adjustment factor, mod(·), sin(·) and cos(·) are the modulus function, sine function and cosine function, respectively, r1 and r2 are random numbers; Step S22: Calculate the fitness value and pre-set the safety distance , calculate the obstacle avoidance penalty factor and path length penalty factor of the candidate global path corresponding to each particle position, and then perform weighted processing according to the weight coefficient to obtain the fitness value of each particle position; Step S23: Design the deviation coefficient, the formula used is as follows: ； In the formula, is the deviation coefficient of the position of the mth particle in the uth search, and are the minimum fitness value and the average fitness value at the u-th search, β is the disturbance amplitude, and u max is the maximum number of searches, is the fitness value of the mth particle position during the uth search; Step S24: Particle position update: The particle position is updated according to the deviation coefficient, and the formula used is as follows: ； In the formula, and are the positions of the mth particle in the u+1th and uth searches respectively, r3, r4 and r5 are random numbers, is a position randomly selected from the M particle positions in the u-th search, δ is the attraction coefficient, is the global optimal position at the u-th search. The global optimal position is the particle position with the minimum fitness value. k1 and k2 are the golden section coefficients. is the deviation coefficient of the position of the mth particle in the uth search; Step S25: Determine the optimal global path; pre-set the fitness threshold, update the fitness value of the particle position, if the fitness value of the global optimal position is lower than the fitness threshold, then the candidate global path represented by the global optimal position is used as the optimal global path, and the intelligent cockpit controls the agricultural machinery to start from the starting point and perform automatic driving according to the optimal global path; otherwise, if the maximum number of searches is reached, return to step S21 to re-initialize the global path; otherwise, return to step S23 to continue searching.

2. The method for controlling the automatic driving of agricultural machinery based on the intelligent cockpit according to claim 1 is characterized in that: In step S4, the local path planning specifically includes the following steps: Step S41: generating a candidate speed combination, including the following steps: Step S411: speed limit; based on the minimum and maximum values ​​of the linear speed v and angular speed ω of the agricultural machine, formulate a first value set of speed combinations ; where v min and v max are the minimum and maximum values ​​of the linear velocity, ω min and ω max are the minimum and maximum angular velocity respectively; Step S412: Minimum turning radius restriction: Based on the minimum turning radius of the agricultural machine, a second value set of speed combinations is formulated ;in, is the minimum turning radius; Step S413: Acceleration and deceleration restrictions: Since the torque of the drive motor is limited, the agricultural machine is constrained when accelerating and decelerating, and a third value set of speed combinations is formulated. ; where v t and ω t are the linear velocity and angular velocity of the agricultural machinery at the current time t, and are the maximum deceleration and acceleration values ​​of the linear velocity, and are the maximum deceleration and acceleration of the angular velocity, is the time interval; Step S414: Braking distance limit; Under the condition of maximum deceleration, a fourth value set of speed combinations is formulated for the agricultural machine to stop before colliding with an obstacle. ;in, is the distance between the trajectory obtained by the agricultural machine traveling at the speed combination (v, ω) and the nearest obstacle; Step S415: Sampling; obtaining the final value set of the agricultural machinery speed combination ,right The linear velocity and angular velocity in the step size are and Take samples and get candidate linear velocity values ​​and candidate angular velocity values, pair the candidate linear velocity values ​​with the candidate angular velocity values, and obtain N v ×N ω candidate speed combinations (v, ω); Step S42: simulate the local path; when the agricultural machine detects a moving obstacle in the observation area V, collect the current position (x, y) and direction angle θ of the agricultural machine, and simulate the local path of the agricultural machine according to each candidate speed combination (v, ω) to obtain N v ×N ω The location of the end of the local path J(v,ω) and direction angle ; Step S43: local path evaluation; Step S44: Determine the optimal local path; from all simulated local paths, select the local path with the largest comprehensive evaluation value as the optimal local path, and the intelligent cockpit controls the agricultural machinery to perform automatic driving according to the optimal local path to avoid moving obstacles.

3. The method for controlling the automatic driving of agricultural machinery based on the intelligent cockpit according to claim 2 is characterized in that: In step S43, the local path evaluation specifically includes the following steps: Step S431: Safety assessment: The safety value of J(v, ω) is obtained by calculating the distance between the end of the local path J(v, ω) and each obstacle in the observation area V. ; The formula used is as follows: ； ； Where E is the number of obstacles detected in the observation area V, γ e is the safety factor of the e-th obstacle, is the distance between the end of the local path J(v,ω) and the e-th obstacle, σ max is the maximum distance error value; Step S432: Power efficiency evaluation: by considering the acceleration at the end of the local path J(v, ω) and the maximum acceleration capability of the agricultural machine, the power efficiency value of J(v, ω) is obtained. ; The formula used is as follows: ； In the formula, and are the linear acceleration and angular acceleration at the end of the local path J(v, ω), respectively; Step S433: Target proximity evaluation: by considering the distance between the end of the local path J(v, ω) and the target point , get the target proximity value of J(v,ω) ; Step S434: Path consistency evaluation; by considering the distance between the end of the local path J(v, ω) and the nearest path point in the global path , and obtain the path consistency value of J(v,ω) ; Step S435: Direction consistency evaluation; by considering the angle difference between the direction angle at the end of the local path J(v, ω) and the direction angle from the agricultural machine to the target point , we get the direction-consistent value of J(v,ω) ; Step S436: Overall evaluation: normalize the safety value, power efficiency value, target proximity value, path consistency value and direction consistency value respectively, and then perform weighted combination to obtain the overall evaluation value of the local path. .

4. The method for controlling the automatic driving of agricultural machinery based on the intelligent cockpit according to claim 1 is characterized in that: In step S1, generating a grid map of the working area specifically includes the following steps: Step S11: collecting images of the working area; collecting images of the working area through sensors installed on the agricultural machinery, and transmitting the collected images of the working area to the intelligent cockpit; Step S12: image preprocessing: in the smart cockpit, the collected working area image is preprocessed, the contrast of the working area image is enhanced by using a histogram equalization algorithm, and the noise in the working area image is removed by using a Gaussian filtering algorithm; Step S13: Image recognition: In the smart cockpit, the boundary data of the working area is identified from the working area image based on the Canny edge detection algorithm, and the obstacle data in the working area image is identified using the YOLO model; Step S14: constructing a grid map; constructing a grid map of the working area in the smart cockpit based on the boundary data and obstacle data identified from the working area image, and calibrating the starting point and the target point.

5. The method for controlling the automatic driving of agricultural machinery based on the intelligent cockpit according to claim 1 is characterized in that: In step S4, the updating of the grid map is to define an observation area V with the agricultural machine as the center and 3 times the safety distance as the radius during the process of the agricultural machine performing automatic driving along the optimal global path. The agricultural machine monitors the environment in the observation area V in real time through sensors and updates the grid map. If a moving obstacle is detected in the observation area V, local path planning is performed; otherwise, the agricultural machine will continue to drive automatically along the optimal global path.

6. The method for controlling the automatic driving of agricultural machinery based on the intelligent cockpit according to claim 1 is characterized in that: In step S5, the automatic driving control is that after the intelligent cockpit controls the agricultural machinery to reach the end of the optimal local path, the grid map is updated in real time. If a moving obstacle is detected again in the observation area V, the local path planning is re-performed; otherwise, the agricultural machinery continues to perform automatic driving along the optimal global path until the agricultural machinery reaches the target point, and the automatic driving ends.

7. An agricultural machinery automatic driving control system based on a smart cockpit, used to implement an agricultural machinery automatic driving control method based on a smart cockpit as claimed in any one of claims 1 to 6, characterized in that: It includes a grid map module for generating a working area, a global path planning module, an update grid map module, a local path planning module and an automatic driving control module; The grid map generating module of the working area collects the working area image and performs preprocessing, edge detection and obstacle recognition in the smart cockpit, constructs the grid map and calibrates the starting point and the target point, and sends the data to the global path planning module; The global path planning module uses the grid map as the search space, combines trigonometric functions with random numbers to dynamically generate particle positions, calculates the fitness value according to the obstacle avoidance penalty factor and the path length penalty factor, updates the particle position according to the deviation coefficient, determines the optimal global path, and the intelligent cockpit controls the agricultural machinery to perform automatic driving according to the optimal global path, and sends the data to the update grid map module; The grid map update module is a module in which the agricultural machine monitors the observation area in real time and updates the grid map, performs local path planning when a moving obstacle is detected, otherwise continues to travel along the optimal global path, and sends data to the local path planning module; The local path planning module formulates a value set based on the value range of linear speed, angular speed, minimum turning radius, acceleration and deceleration limits, and braking distance limits, and then samples to obtain candidate speed combinations, simulates local paths under different candidate speed combinations, and comprehensively evaluates the local paths from five aspects: safety, power efficiency, target proximity, path consistency, and direction consistency, selects the optimal local path with the maximum comprehensive evaluation value, and sends the data to the automatic driving control module; The automatic driving control module is an intelligent cockpit that controls the agricultural machinery to travel along the optimal local path and updates the grid map in real time. If a moving obstacle is detected, the local path is replanned, otherwise it continues to travel along the optimal global path until it reaches the target point.

Images