A tractor-UAV cooperative operation control method for irregular fields

By using a tractor-drone collaborative operation control method, the problem of low matching degree of agricultural machinery operation trajectory in irregular fields was solved, achieving efficient field coverage and drone blind spot filling, and improving operation coverage and stability.

CN123113984APending Publication Date: 2026-11-17SOUTH CHINA AGRICULTURAL UNIVERSITY
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202611460312.X
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2026-09-20
Publication Date
2026-11-17

AI Technical Summary

Technical Problem

Traditional path planning algorithms struggle to accurately identify complex field boundary shapes in irregular fields, resulting in low matching between agricultural machinery operation trajectories and field boundaries. This leads to missed tillage, overlapping tillage, or inaccessible coverage blind spots, reducing the overall field coverage rate and increasing the cost of ineffective operations.

Method used

A tractor-UAV collaborative operation control method is adopted. By constructing an operable area in an irregular field, discretizing the coverage unit, the tractor-accessible coverage area and the UAV blind spot coverage area are generated. Combined with the air-ground task allocation cost function and a unified coordinate system, the collaborative operation path planning of tractors and UAVs and the UAV blind spot coverage trajectory tracking are realized.

Benefits of technology

It improved the operational coverage of irregular fields, reduced the costs of missed, overlapping, and waiting, enhanced the stability and coverage of air operations, and enabled drones to take off and land at fixed points on tractor-mounted platforms.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN123113984A_ABST
    Figure CN123113984A_ABST
Patent Text Reader

Abstract

This invention discloses a tractor-UAV collaborative operation control method for irregular fields, comprising: S1, constructing a workable area for the irregular field; S2, discretizing the workable area into coverage units to generate a tractor-accessible coverage area, and extracting areas difficult for the tractor to cover and residual coverage units as UAV blind spot coverage areas; S3, constructing an air-to-ground task allocation cost function to generate a tractor full-coverage operation path and a UAV blind spot coverage trajectory; S4, unifying the tractor, UAV, field coverage units, and take-off and landing platform into the same coordinate system through rigid body coordinate transformation; S5, performing adaptive Kalman fusion by combining positioning and attitude information; S6, the tractor and UAV collaboratively complete the full-coverage operation of the irregular field according to the task sequence. This invention can improve the sowing coverage rate of irregular fields in hilly areas, and reduce missed planting, overlapping operations, and inaccessible coverage blind spots.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This invention belongs to the field of path planning technology, specifically relating to a tractor-drone collaborative operation control method for irregular fields. Background Technology

[0002] Most domestic and international research on full-coverage traversal algorithms focuses on regularly shaped fields, such as rectangular or regular polygonal fields. In these scenarios, agricultural machinery can typically achieve high coverage rates by using parallel reciprocating paths, ox-plowing paths, or regular strip decomposition paths.

[0003] Irregular fields typically feature polygonal boundaries, notches, sharp corners, narrow passages, sloping boundaries, and internal obstacles. When traditional path planning algorithms are applied to irregular fields, they often struggle to accurately identify complex field boundary shapes and cannot dynamically adapt to changes in field contours. This results in low matching between agricultural machinery trajectories and field boundaries, leading to problems such as missed tillage in corner areas, overlapping tillage, or inaccessible coverage blind spots. Consequently, the overall field coverage rate is reduced, and the cost of ineffective operations increases.

[0004] To address the aforementioned issues, incorporating drones into irregular field operations allows tractors to handle efficient ground work in accessible areas, while drones fill in blind spots in corners, recesses, areas with obstacles, and areas inaccessible to tractors. This creates an air-ground collaborative operation system, promoting intelligent, precise, high-quality, and efficient agricultural machinery operations. However, achieving full air-ground collaborative coverage is not simply a matter of superimposing tractor paths and drone flight tracks. It also requires solving problems such as modeling the coverage area of ​​irregular fields, determining tractor-accessible areas, extracting blind spot areas for drones, dynamic task scheduling, timing matching of air and ground operations, coordinated action response, and precise take-off and landing of drones on tractor-mounted platforms. Summary of the Invention

[0005] The main objective of this invention is to overcome the shortcomings and deficiencies of the prior art and to propose a tractor-drone collaborative operation control method for irregular fields.

[0006] To achieve the above objectives, the present invention adopts the following technical solution:

[0007] A method for collaborative operation control of tractors and unmanned aerial vehicles (UAVs) in irregular fields includes the following steps:

[0008] S1. Construct workable areas for irregularly shaped fields;

[0009] S2. Discretize the coverage units of the workable area, generate the coverage area that can be reached by the tractor based on the tractor's passage constraints and turning constraints, and extract the areas that are difficult for the tractor to cover and the remaining coverage units as the blind spot coverage area of ​​the UAV.

[0010] S3. Construct an air-to-ground task allocation cost function. Based on coverage benefits, transfer distance, overlap penalty, operation timing waiting cost, and energy consumption cost, allocate coverage units to tractors or drones, and generate full coverage operation paths for tractors and blind spot filling flight paths for drones.

[0011] S4. Establish a global coordinate system, a tractor body coordinate system, an unmanned aerial vehicle (UAV) system, and a take-off and landing platform coordinate system. Through rigid body coordinate transformation, unify the tractor, UAV, field coverage unit, and take-off and landing platform into the same coordinate system.

[0012] S5. Obtain the distance and azimuth angle of the UAV relative to the tractor, and combine the positioning information and attitude information to perform adaptive Kalman fusion to obtain the estimated values ​​of the position, speed and attitude of the UAV relative to the tractor.

[0013] S6. The UAV adopts a dynamic forward-looking L1 guidance tracking trajectory with integrated positioning reliability. The tractor uses LQR control law to generate the relative position correction of the UAV based on the relative state between the air and the ground, and uses PID control law to adjust the flight altitude, relative distance or operation execution amount, so that the tractor and UAV can work together to complete the full coverage operation of irregular fields according to the task sequence.

[0014] Compared with the prior art, the present invention has the following advantages and beneficial effects:

[0015] 1. The method of this invention establishes workable area and blind spot filling area models for irregular field boundary notches, obstacles and narrow passages, improving the adaptability to complex fields in hilly areas; reduces the costs of missed work, overlap and waiting by using an air-ground task allocation cost function; improves the stability of air-ground relative state estimation by using unified coordinate transformation and adaptive Kalman fusion; realizes UAV blind spot filling trajectory tracking, relative position holding and operation state adjustment by using dynamic forward look-ahead L1, LQR and PID hierarchical control with fixed position confidence feedback; realizes fixed-point take-off and landing of UAV on tractor carrier by initial relative coordinate reuse, dynamic take-off and landing point prediction, terminal LQR correction and variable gain PID vertical control. Attached Figure Description

[0016] Figure 1 This is a flowchart of the method of the present invention.

[0017] Figure 2 This is a schematic diagram of a scenario involving coordinated full-coverage sowing operations on irregularly shaped open fields, as illustrated in the embodiment. Detailed Implementation

[0018] The present invention will be further described in detail below with reference to the embodiments and accompanying drawings, but the embodiments of the present invention are not limited thereto.

[0019] Examples; such as Figure 1 and Figure 2 As shown, a tractor-UAV collaborative operation control method for irregular fields includes the following steps:

[0020] S1. Constructing workable areas for irregularly shaped fields; specifically:

[0021] Obtain the boundaries of irregular fields, the boundaries of internal obstacles, the operating width of tractors, the operating width of drones, the minimum turning radius of tractors, and the safe buffer distance to construct the workable area of ​​irregular fields;

[0022] The workable area of ​​an irregular field is obtained by subtracting the internal obstacle polygons and the safety buffer zone from the outer boundary polygon, and is represented as follows:

[0023] ;

[0024] ;

[0025] ;

[0026] in, The outer boundary of the irregular field is a polygon. Let M be the polygon containing the m-th obstacle inside the polygon, and M be the number of obstacles. For safety buffer structure elements; The area enclosed by the outer boundary of the field. The union of all internal obstacle regions. This indicates a morphological expansion or safety buffer operation. Because it belongs to The workable area is formed by points that are not part of the obstacle safety buffer zone;

[0027] S2. Discretize the coverage units of the workable area, generate the coverage area that can be reached by the tractor based on the tractor's passage constraints and turning constraints, and extract the areas that are difficult for the tractor to cover and the remaining coverage units as the blind spot coverage area of ​​the UAV.

[0028] In this embodiment, the workable area is discretized into coverage units, specifically by dividing the workable area of ​​irregular fields. By resolution Discretize the set of coverage cells. A coverage cell whose center point is located inside the workable area is denoted as an effective coverage cell. The set of effective coverage cells is represented as follows:

[0029] ;

[0030] in, For the center of the i-th coverage unit, This represents the actual area corresponding to the i-th coverage unit. This represents the total number of effective coverage units; each coverage unit corresponds to one working area. If a square grid is used, then If a boundary clipping unit is used, then for and Area of ​​intersection; subsequent and All indicate that they are composed of several The region formed by the union, rather than a set of points consisting only of the central point;

[0031] In step S2, the tractor-accessible coverage area consists of coverage units that meet the conditions of working width, minimum turning radius, slope, obstacle safety distance, and boundary indentation; the UAV blind spot coverage area is the residual coverage units after deducting the tractor-accessible coverage area from the effective coverage unit set, and includes the notch boundary neighborhood, narrow passage area, and obstacle buffer neighborhood.

[0032] The specific determination of the tractor's reachable coverage area is as follows:

[0033] According to the working width of the tractor Minimum turning radius Slope threshold and boundary safety distance First, construct the tractor reachability constraint margin vector, and then determine whether the covering unit can be operated by the tractor based on the margin vector:

[0034] ;

[0035] ;

[0036] in, Let be the reachability constraint margin vector of the tractor in the i-th covering unit; This indicates the distance from the coverage unit to the boundary of the workable area. Indicates the width of the local channel where the coverage unit is located. Indicates the feasible radius for a local turn. Indicates the local slope. express The minimum value among all components; when this minimum value is not less than zero. Select 1 if the value is 1, otherwise select 0. Based on... and The coverage areas accessible by tractors and the coverage areas filled by drones are obtained:

[0037] ;

[0038] ;

[0039] For irregularly shaped fields This typically addresses areas such as notches, sharp corners, areas near obstacles, narrow passages, boundary buffer zones, and areas where tractors have difficulty turning around. The introduction of drone-based blind spot detection eliminates the need for tractors to force their way into these areas, thereby reducing mechanical damage and the cost of ineffective turns.

[0040] S3. Construct an air-to-ground task allocation cost function. Based on coverage benefits, transfer distance, overlap penalty, operation timing waiting cost, and energy consumption cost, allocate coverage units to tractors or drones, and generate full coverage operation paths for tractors and blind spot filling flight paths for drones.

[0041] In this embodiment, the air-to-ground task allocation cost function includes coverage gain, transfer distance, overlap penalty, energy consumption cost, and waiting cost, with coverage unit allocation variables. satisfy:

[0042] ;

[0043] ;

[0044] ;

[0045] in, This indicates that the work is carried out by a tractor. This indicates that the operation is carried out by a drone; m can be either T or U, representing a tractor and a drone, respectively. From the current task point to the coverage unit The cost of transition distance, For the cost of energy consumption, Penalties for overlap with already planned work areas, The cost of waiting in empty spaces or time-series conflicts. To cover revenue, to These are the weighting coefficients;

[0046] Based on the air-to-ground task allocation cost function, the goal of task allocation is to minimize the overall cost while ensuring that all effective coverage units are covered.

[0047] ;

[0048] ;

[0049] in, To allow for safe overlapping sets, and These represent the actual operational footprint areas formed by the tractor and the drone according to the planned trajectory, respectively.

[0050] Based on the coverage unit allocation variables, tractor coverage sets and drone coverage sets are obtained, and coverage rate, overlap rate, and missed coverage rate are used as evaluation indicators for full coverage:

[0051] ;

[0052] ;

[0053] ;

[0054] in, Indicates the area measurement of a region. The coverage rate is calculated from the assigned area. Based on actual operational footprint and Calculated overlap rate This represents the omission rate.

[0055] S4. Establish a global coordinate system, a tractor body coordinate system, an unmanned aerial vehicle (UAV) system, and a take-off and landing platform coordinate system. Through rigid body coordinate transformation, unify the tractor, UAV, field coverage unit, and take-off and landing platform into the same coordinate system.

[0056] In this embodiment, let the position of the tractor in the global coordinate system be... The attitude angle is , and The rotation matrix from the tractor body coordinate system T to the global coordinate system G is:

[0057] ;

[0058] For any point in the tractor body coordinate system Its global coordinates are:

[0059] ;

[0060] Let the global position, velocity, and attitude matrices of the UAV be respectively... , and The tractor's global speed is The position, velocity, and attitude of the UAV relative to the tractor are unified to the tractor's coordinate system using the following formula:

[0061] ;

[0062] ;

[0063] ;

[0064] in, This indicates the matrix transpose.

[0065] S5. Obtain the distance and azimuth angle of the UAV relative to the tractor, and combine the positioning information and attitude information to perform adaptive Kalman fusion to obtain the estimated values ​​of the position, speed and attitude of the UAV relative to the tractor.

[0066] In this embodiment, a UWB reference module mounted on the tractor and a UWB tag module mounted on the drone are used to obtain the distance and azimuth of the drone relative to the tractor; the UWB reference module and the UWB tag module output the relative distance. and relative azimuth Let the relative position of the UAV in the tractor's coordinate system be... The UWB measurement model is then:

[0067] ;

[0068] ;

[0069] ;

[0070] because This is the three-dimensional slant distance, and For the corresponding horizontal azimuth angle, first from Calculate horizontal distance with vertical relative component UWB relative position observations are constructed based on relative distance and relative azimuth:

[0071] ;

[0072] ;

[0073] ;

[0074] ;

[0075] ;

[0076] in, The three-dimensional relative slope distance is measured by single-reference, single-label UWB. This is the horizontal relative distance converted from slant distance. The relative azimuth angle. and These are ranging noise and azimuth noise, respectively. , and For the position components of the UAV relative to the tractor body coordinate system, these are the three axes of the UAV's position. and These are the longitudinal and lateral relative position observations of the UWB structure, respectively; It is obtained from height measurement, positioning height difference, or filtered prediction values; This is the UWB ranging residual, used for takeoff clearance and anomaly measurement judgment.

[0077] To describe the adaptive Kalman fusion process in detail, the state vector is defined as follows:

[0078] ;

[0079] in, , , These are the positional components of the drone relative to the tractor. , , These are the relative velocity components. Relative heading angle The sensor has zero bias. Based on the assumption of uniform short-time motion, the position and velocity sub-states satisfy:

[0080] ;

[0081] ;

[0082] Therefore, the state prediction model can be written as:

[0083] ;

[0084] in, The state transition matrix is ​​composed of the position-velocity submatrix. Heading submatrix and zero-bias matrix composition:

[0085] ;

[0086] ;

[0087] ;

[0088] in, This is a six-dimensional state transition submatrix corresponding to relative position and relative velocity. For the relative heading angle state transition term, This is the zero-bias state transition term for the sensor. It is a third-order identity matrix. It is a third-order zero matrix. The sampling period is [period]. During calculation, the upper half... The second half represents the integral contribution of the relative velocity at the previous moment to the current position. This indicates that the relative velocity within a short sampling period is propagated from the previous moment and corrected by process noise.

[0089] The measurement vector consists of localization relative position, UWB relative position observation, and attitude and flight vector measurement:

[0090] ;

[0091] ;

[0092] In the implementation methods of locating relative position, constructing relative position using UWB, and directly corresponding state quantities for navigation vector measurements, the Jacobian matrix is ​​measured. Expanded by measurement source:

[0093] ;

[0094] Process noise covariance matrix Expanding the state vector element by element as follows:

[0095] ;

[0096] in, , , The standard deviation of the relative position process noise. , , The standard deviation of the noise in the relative velocity process. The standard deviation of noise during the relative heading process. The standard deviation of the zero-bias process noise is given.

[0097] Measurement noise covariance matrix Expanding the measurement vector element by element as follows:

[0098] ;

[0099] in, To measure the standard deviation of noise for planar positioning, To measure the standard deviation of noise for height positioning, , , These represent the standard deviations of the longitudinal, lateral, and vertical relative position observation noise of the UWB structure, respectively. The standard deviation of the attitude and heading vector measurement noise. Matrix Each diagonal element corresponds to a specific measurement source, and the Kalman gain is calculated using... Determine the weight of different measurements for state updates.

[0100] The prediction error covariance matrix is ​​obtained by propagating the previous time step update covariance, the state transition matrix, and the process noise covariance, and its matrix dimension is consistent with the eight-dimensional state vector.

[0101] ;

[0102] ;

[0103] ;

[0104] For measurement equations Linearization at the predicted state yields the measurement Jacobian matrix. And update according to the following formula:

[0105] ;

[0106] ;

[0107] ;

[0108] in, For the prediction error covariance matrix, The updated error covariance matrix, This is the Kalman gain. To adapt to the effects of hilly terrain shading, vibration, and low-altitude multipath propagation, residual adaptive noise adjustment is introduced:

[0109] ;

[0110] ;

[0111] in, To measure the residual, The initial measurement noise covariance matrix, This is the adaptive adjustment coefficient.

[0112] S6. The UAV adopts a dynamic forward-looking L1 guidance tracking trajectory with integrated positioning reliability. The tractor uses LQR control law to generate the relative position correction of the UAV based on the relative state between the air and the ground, and uses PID control law to adjust the flight altitude, relative distance or operation execution amount, so that the tractor and UAV can work together to complete the full coverage operation of irregular fields according to the task sequence.

[0113] In this embodiment, the UAV employs a dynamic forward-looking L1 guidance tracking trajectory that integrates positioning confidence. Let the forward-looking distance selected by the UAV on the trajectory be... The target point, the angle between the velocity direction and the line connecting the target point is... Then the lateral acceleration is:

[0114] ;

[0115] ;

[0116] ;

[0117] in, For guidance speed with low-speed protection, The minimum speed required to participate in guidance calculations; when the drone's speed is below At that time, adopt As a denominator, it avoids singular amplification of heading angular velocity commands at low speeds.

[0118] To incorporate the idea of ​​using positional confidence in control and to avoid tracking lag or oscillations caused by fixed forward look-ahead distance in notches, narrow channels, and obstacle neighborhoods, this invention further feeds back the covariance of the adaptive Kalman fusion output to the L1 guidance layer. Let... To integrate the submatrices in the covariance matrix that correspond to the relative positional state, a positioning uncertainty index is defined. and location reliability factor :

[0119] ;

[0120] ;

[0121] in, This is the three-dimensional submatrix in the Kalman covariance matrix corresponding to the relative position state. Calculated using the matrix F norm, a larger value indicates greater uncertainty in relative positioning; As a reference scale for locating uncertainty, it is used to make It becomes a dimensionless quantity; The positioning uncertainty is converted into a dimensionless confidence factor, the value of which varies with... Increases and decreases.

[0122] Based on the UAV's speed, lateral deviation, and positioning reliability, the unlimited and limited values ​​for the dynamic forward-looking distance are as follows:

[0123] ;

[0124] ;

[0125] in, This refers to the unlimited dynamic forward look-ahead distance. Based on the basic forward sight distance, For the drone's horizontal speed, The lateral deviation of the UAV relative to the blind spot fill track. , and This is a non-negative adjustment coefficient. In the calculation, the velocity term is used to avoid oscillations caused by excessively short forward sight distances at high speeds, the lateral deviation term is used to increase the tracking margin when deviating from the track, and the confidence term is used to increase the forward sight distance when positioning is uncertain.

[0126] Furthermore, the dynamic forward look-ahead is obtained by superimposing the forward look-ahead of the blind spot fill track, the expected relative position of the tractor, and the cooperative correction amount:

[0127] ;

[0128] in, This represents the parameterization function for the blind spot fill track. This is the current progress of the flight path. To advance along the gap-filling flight path The geometric foreground point at the distance; To project the desired relative position onto the global coordinate system, This is the relative position correction value generated for the LQR collaborative layer. The calculation first determines the forward look-ahead point of the flight path, then superimposes the expected relative position of the tractor-UAV and the correction value to obtain the dynamic target point actually tracked by the UAV.

[0129] Considering the impact of the tractor's motion changes on the forward-looking target of the UAV, velocity feedforward compensation along the path normal is introduced to obtain the dynamic L1 lateral acceleration:

[0130] .

[0131] in, The angle between the direction of the UAV's velocity and the direction of the line connecting the dynamic forward-looking point. To compensate for the unit normal vector of the blind track at the current progress, and These represent the velocities of the tractor and the drone in the global coordinate system, respectively. This represents the velocity feedforward coefficient.

[0132] In this embodiment, at the air-ground cooperative control layer, the tractor sets the desired relative position of the UAV according to the task sequence. and expected relative speed Construct the cooperative error state:

[0133] ;

[0134] in, This is a six-dimensional cooperative error state vector, where the first three dimensions represent the actual relative position of the UAV. Relative position to expectations The positional error between them, the latter three dimensions are the actual relative velocities. Relative speed with respect to expectations Speed ​​error between;

[0135] The cooperative error state is described by a second-order relative motion model, and its state equation and matrix expansion are as follows:

[0136] ;

[0137] ;

[0138] in, The six-dimensional cooperative error state matrix, To control the input matrix, For the relative position or speed correction control of the UAV.

[0139] Let the performance index of the quadratic form be:

[0140] ;

[0141] ;

[0142] in, For collaborative control performance indicators, The sixth-order cooperative error weight matrix is... The third-order control input weight matrix; , and These are the weights for the three relative position errors. , and These are the weights for the three relative velocity errors. , and These are the three correction input weights. Let S be a positive definite solution to the algebraic Riccati equation, then S and the feedback gain satisfy:

[0143] ;

[0144] ;

[0145] in, and They are respectively and The transpose of S, where S is the transpose of S. and The solution matrix of the Riccati equations is determined; This is the LQR feedback gain matrix. The calculation is first based on... , , and Solve for S, then obtain .

[0146] The LQR feedback gain is obtained by solving the problem. Then, the correction control quantity is:

[0147] ;

[0148] in, This is the correction control quantity output by the LQR collaborative layer.

[0149] In this embodiment, a PID control law is used to adjust the flight altitude, relative distance, or workload, specifically as follows:

[0150] A PID control law with amplitude limiting and anti-integral saturation is used to adjust the UAV's flight altitude, relative distance, and task execution. The PID control law is expressed as:

[0151] ;

[0152] ;

[0153] Where j represents the type of slow variable being adjusted. For the corresponding error, , and These are the proportional, integral, and derivative gains, respectively. The integral term is used to eliminate steady-state error, and the derivative term is used to suppress excessively rapid changes in error. This is the execution instruction after bandwidth limiting. and These are the minimum and maximum control limits allowed by the actuator, respectively.

[0154] In addition, during the takeoff or landing phase of the UAV, the tractor end makes a takeoff clearance judgment based on positioning uncertainty, UWB residual and attitude error, and reuses the initial relative coordinates calibrated before takeoff to calculate the target point of the mobile takeoff and landing platform. The UAV performs phased closed-loop control based on the target takeoff and landing point and attitude error to realize the fixed-point takeoff and landing of the UAV on the tractor carrier.

[0155] The initial relative coordinates, the target point of the moving takeoff and landing platform, the takeoff clearance determination, the delay compensation target point, the landing relative error, the attitude error, and the terminal takeoff and landing control variables are as follows:

[0156] ;

[0157] ;

[0158] ;

[0159] ;

[0160] ;

[0161] ;

[0162] ;

[0163] ;

[0164] ;

[0165] ;

[0166] ;

[0167] in, and The global coordinates of the tractor and the drone were used to determine the takeoff time; The target offset of the center of the take-off and landing platform in the tractor body coordinate system. The fixed installation offset from the launch beacon center to the platform center; For takeoff safety normalized vector, This is a takeoff clearance sign. To locate uncertainty indicators, For UWB ranging residuals, Let be the attitude error vector between the UAV and the platform. , and These are the positioning uncertainty, UWB residual, and attitude error threshold, respectively. This refers to the position of the takeoff and landing platform center in the global coordinate system. The global position of the tractor positioning reference point; This is the projection of the platform offset onto the global coordinate system. This refers to the landing error in the platform coordinate system. This is the rotation matrix from the platform coordinate system to the global coordinate system; , and This represents the positional error in the three directions within the platform coordinate system. , and For relative speed error, This refers to the relative heading error; This is a seven-dimensional terminal landing error state vector. This is the seventh-order terminal state error weight matrix. This is the fourth-order end-control input weight matrix; , and These represent the position error weights in the three directions of the platform coordinate system. , and These are the relative velocity error weights, For heading error weighting, , , and These are the four end-point correction input weights; For vertical control, This represents the proportional gain that varies with positioning uncertainty. and These are the vertical integral gain and the differential gain, respectively. This represents the vertical position error.

[0168] It should also be noted that, in this specification, terms such as "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0169] The above description of the disclosed embodiments enables those skilled in the art to make or use the invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the invention. Therefore, the invention is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A tractor-UAV collaborative operation control method for irregular field plots, characterized in that, Includes the following steps: S1. Construct workable areas for irregularly shaped fields; S2. Discretize the coverage units of the workable area, generate the coverage area that can be reached by the tractor based on the tractor's passage constraints and turning constraints, and extract the areas that are difficult for the tractor to cover and the remaining coverage units as the blind spot coverage area of ​​the UAV. S3. Construct an air-to-ground task allocation cost function. Based on coverage benefits, transfer distance, overlap penalty, operation timing waiting cost, and energy consumption cost, allocate coverage units to tractors or drones, and generate full coverage operation paths for tractors and blind spot filling flight paths for drones. S4. Establish a global coordinate system, a tractor body coordinate system, an unmanned aerial vehicle (UAV) system, and a take-off and landing platform coordinate system. Through rigid body coordinate transformation, unify the tractor, UAV, field coverage unit, and take-off and landing platform into the same coordinate system. S5. Obtain the distance and azimuth angle of the UAV relative to the tractor, and combine the positioning information and attitude information to perform adaptive Kalman fusion to obtain the estimated values ​​of the position, speed and attitude of the UAV relative to the tractor. S6. The UAV adopts a dynamic forward-looking L1 guidance tracking trajectory with integrated positioning reliability. The tractor uses LQR control law to generate the relative position correction of the UAV based on the relative state between the air and the ground, and uses PID control law to adjust the flight altitude, relative distance or operation execution amount, so that the tractor and UAV can work together to complete the full coverage operation of irregular fields according to the task sequence.

2. The method for collaborative operation control of tractors and unmanned aerial vehicles (UAVs) in irregular fields according to claim 1, characterized in that, Step S1 is as follows: Obtain the boundaries of irregular fields, the boundaries of internal obstacles, the operating width of tractors, the operating width of drones, the minimum turning radius of tractors, and the safe buffer distance to construct the workable area of ​​irregular fields; The workable area of ​​an irregular field is obtained by subtracting the internal obstacle polygons and the safety buffer zone from the outer boundary polygon, and is represented as follows: ; ; ; in, The outer boundary of the irregular field is a polygon. Let M be the polygon containing the m-th obstacle inside the polygon, where M is the number of obstacles. For safety buffer structure elements; The area enclosed by the outer boundary of the field. The union of all internal obstacle regions. This indicates a morphological expansion or safety buffer operation. Because it belongs to The workable area is defined by points that are not within the safety buffer zone of obstacles; In step S2, the workable area is discretized into coverage units, specifically by dividing the workable area of ​​irregular fields. By resolution Discretize the set of coverage cells. A coverage cell whose center point is located inside the workable area is denoted as an effective coverage cell. The set of effective coverage cells is represented as follows: ; in, For the center of the i-th coverage unit, This represents the actual area corresponding to the i-th coverage unit. This represents the total number of effective coverage units; each coverage unit corresponds to one working area. If a square grid is used, then If a boundary clipping unit is used, then for and Area of ​​intersection; subsequent and All indicate that they are composed of several The region formed by the union, rather than a set of points consisting only of the central point; In step S2, the tractor-accessible coverage area consists of coverage units that meet the conditions of working width, minimum turning radius, slope, obstacle safety distance, and boundary indentation; the UAV blind spot coverage area is the residual coverage units after deducting the tractor-accessible coverage area from the effective coverage unit set, and includes the notch boundary neighborhood, narrow passage area, and obstacle buffer neighborhood.

3. The method for collaborative operation control of tractors and unmanned aerial vehicles (UAVs) in irregular fields according to claim 2, characterized in that, In step S3, the cost function for allocating air-to-ground tasks includes coverage gain, transfer distance, overlap penalty, energy cost, and waiting cost, and the coverage unit allocation variable... satisfy: ; ; ; in, This indicates that the work is carried out by a tractor. This indicates that the operation is carried out by a drone; m can be either T or U, representing a tractor and a drone, respectively. From the current task point to the coverage unit The cost of transition distance, For the cost of energy consumption, Penalties for overlap with already planned work areas, The cost of waiting in empty spaces or time-series conflicts. To cover revenue, to These are the weighting coefficients; Based on the air-to-ground task allocation cost function, the goal of task allocation is to minimize the overall cost while ensuring that all effective coverage units are covered. ; ; in, To allow for safe overlapping sets, and These represent the actual operational footprint areas formed by the tractor and the drone according to the planned trajectory, respectively. Based on the coverage unit allocation variables, tractor coverage sets and drone coverage sets are obtained, and coverage rate, overlap rate, and missed coverage rate are used as full coverage evaluation indicators: ; ; ; in, Indicates the area measurement of a region. The coverage rate is calculated from the assigned area. Based on actual operational footprint and Calculated overlap rate This represents the omission rate.

4. The method for collaborative operation control of tractors and unmanned aerial vehicles (UAVs) in irregular fields according to claim 1, characterized in that, In step S4, let the position of the tractor in the global coordinate system be... The attitude angle is , and The rotation matrix from the tractor body coordinate system T to the global coordinate system G and the coordinate transformation of any point satisfy: ; ; Let the global position, velocity, and attitude matrices of the UAV be respectively... , and The tractor's global speed is The position, velocity, and attitude of the UAV relative to the tractor are unified to the tractor's coordinate system using the following formula: ; ; 。 5. The method for collaborative operation control of tractors and unmanned aerial vehicles (UAVs) in irregular fields according to claim 1, characterized in that, In step S5, a UWB reference module installed on the tractor and a UWB tag module installed on the drone are used to obtain the distance and azimuth angle of the drone relative to the tractor; the UWB reference module and the UWB tag module output the relative distance. and relative azimuth Let the relative position of the UAV in the tractor's coordinate system be... The UWB measurement model is then: ; ; ; Construct UWB relative position observations based on relative distance and relative azimuth: ; ; ; ; ; in, The three-dimensional relative slope distance is measured by single-reference, single-label UWB. This is the horizontal relative distance converted from slant distance. The relative azimuth angle. and These are ranging noise and azimuth noise, respectively. , and For the position components of the UAV relative to the tractor body coordinate system, these are the three axes of the UAV's position. and These are the longitudinal and lateral relative position observations of the UWB structure, respectively; It is obtained from height measurement, positioning height difference, or filtered prediction values; This is the UWB ranging residual, used for takeoff clearance and anomaly measurement judgment.

6. The method for collaborative operation control of tractors and unmanned aerial vehicles (UAVs) in irregular fields according to claim 1, characterized in that, In step S6, the UAV adopts a dynamic forward-looking L1 guidance tracking trajectory with fused positioning confidence. The positioning uncertainty index, positioning confidence factor, dynamic forward-looking distance, forward-looking point, and lateral acceleration satisfy the following: ; ; ; ; ; ; ; ; 。 7. The method for collaborative operation control of tractors and unmanned aerial vehicles (UAVs) in irregular fields according to claim 1, characterized in that, In step S6, the tractor end constructs a cooperative error state and solves the LQR feedback control law based on the quadratic performance index: ; ; ; ; in, This is a six-dimensional cooperative error state vector, where the first three dimensions represent the actual relative position of the UAV. Relative position to expectations The positional error between them, the latter three dimensions are the actual relative velocities. Relative speed with respect to expectations Speed ​​error between; For collaborative control performance indicators, The sixth-order cooperative error weight matrix is... The third-order control input weight matrix; , and These are the weights for the three relative position errors. , and These are the weights for the three relative velocity errors. , and These are the three correction input weights; This is the LQR feedback gain matrix; This is the correction control quantity output by the LQR collaborative layer.

8. The method for collaborative operation control of tractors and unmanned aerial vehicles (UAVs) in irregular fields according to claim 1, characterized in that, In step S6, a PID control law is used to adjust the flight altitude, relative distance, or operational quantity, specifically as follows: A PID control law with amplitude limiting and anti-integral saturation is used to adjust the UAV's flight altitude, relative distance, and task execution. The PID control law is expressed as: ; ; Where j represents the type of slow variable being adjusted. For the corresponding error, , and These are the proportional, integral, and derivative gains, respectively. The integral term is used to eliminate steady-state error, and the derivative term is used to suppress excessively rapid changes in error. This is the execution instruction after bandwidth limiting. and These are the minimum and maximum control limits allowed by the actuator, respectively.