Method and system for controlling operation of a crop harvester

Abstract

The application provides a crop harvester operation control method and system. The method comprises the following steps: based on a preset dynamic mapping model of density, terrain and pose, mapping density distribution data and terrain feature data into target pose parameters of a controlled object; acquiring actual pose parameters of the controlled object, inputting the actual pose parameters and the target pose into an adaptive controller, and dynamically calculating a pose error compensation amount through a closed-loop control algorithm; generating a pose control signal based on the pose error compensation amount, driving a pose adjusting structure to calibrate the pose of the controlled object, so as to adjust the working height and inclination angle of an end effector installed at the front end of a multi-degree-of-freedom actuator. The technical scheme provided by the application can automatically and accurately adjust the height and inclination angle of the end effector in real time according to the crop density and terrain undulation, and keep the optimal harvesting pose.

Description

Technical Field

[0001] This application relates to the field of agricultural machinery technology, and in particular to an operation control method and system for a crop harvester. Background Technology

[0002] Agricultural harvesting machines are widely used in modern agricultural production, significantly improving operational efficiency and crop quality. Taking corn harvesters, wheat harvesters, and leek harvesters as examples, the working posture of the end effector of these machines is crucial to harvesting quality and efficiency. Therefore, optimizing cutting precision and reducing crop damage rates to improve harvesting efficiency and crop yield quality provides technical support for large-scale planting.

[0003] In existing technologies, the operation control process of crop harvesters is typically simplified, relying mainly on preset parameters or basic sensor feedback. The specific process includes: first, roughly acquiring crop height data using a single sensor, such as an ultrasonic or infrared device; second, comparing the actual height with a preset target value based on a fixed algorithm, such as a proportional-integral-derivative controller, generating a height deviation signal; and then, driving the actuator to perform linear adjustments to regulate the working height and tilt angle of the end effector. Throughout this process, the operation control of existing solutions relies on empirical settings and cannot dynamically respond to changes in the field. While this process can achieve basic adjustments, it lacks the ability to proactively adapt to complex environmental factors.

[0004] Therefore, the main drawback of existing technologies lies in their inability to effectively address field non-uniformity, such as differences in crop density distribution or terrain undulations, which can easily lead to inaccurate control of harvesting height and tilt angle. Specifically, the height deviation signal is biased, and the disturbance compensation is insufficient, resulting in large fluctuations in the end effector's pose and uneven cutting, increasing the risk of crop damage, such as root tearing or leaf breakage. At the same time, the fixed gain control algorithm has poor stability and slow response under multi-parameter disturbances, affecting harvesting quality and efficiency. Summary of the Invention

[0005] This application provides a method and system for controlling the operation of a crop harvester, which solves the problem of inaccurate control of harvesting height and tilt angle in the prior art, and the inability to dynamically respond to changes in the field.

[0006] In a first aspect, this application provides an operation control method for a crop harvester, including:

[0007] Acquire density distribution data of the target object and terrain feature data of the area where the target object is located;

[0008] Based on a preset dynamic mapping model of density, terrain and pose, the density distribution data and the terrain feature data are mapped to the target pose parameters of the controlled object. The target pose parameters include: height setting value and tilt angle setting value. The controlled object is a multi-degree-of-freedom actuator of a crop harvester.

[0009] The actual pose parameters of the controlled object are obtained, and the actual pose parameters and the target pose are input into the adaptive controller. The pose error compensation amount is dynamically calculated through the closed-loop control algorithm.

[0010] Based on the posture error compensation amount, a posture control signal is generated to drive the posture adjustment structure to perform posture calibration on the controlled object, so as to adjust the working height and tilt angle of the end effector installed at the front end of the multi-degree-of-freedom actuator.

[0011] Optionally, the step of inputting the actual pose parameters and the target pose into the adaptive controller, and dynamically calculating the pose error compensation amount through a closed-loop control algorithm, includes:

[0012] In the adaptive controller, the actual height value and the actual tilt angle value are obtained from the actual pose parameters;

[0013] The set height value and the actual height value are tracked at a set point to generate a height deviation signal; the set tilt angle value and the actual tilt angle value are tracked at the set point to generate a tilt angle deviation signal.

[0014] By using the multi-parameter coupling compensation module of the adaptive controller and combining it with the dynamic transfer function of the multi-degree-of-freedom actuator, disturbance estimation is performed on the height deviation signal and the tilt angle deviation signal to obtain a pose disturbance component that includes the height disturbance component and the tilt angle disturbance component.

[0015] The height disturbance component and the tilt disturbance component are compared with the corresponding preset stability boundary conditions to obtain the dynamically adjusted height channel gain coefficient and tilt channel gain coefficient.

[0016] Using a closed-loop control algorithm, the height deviation signal and the tilt deviation signal are processed based on the height channel gain coefficient and the tilt channel gain coefficient to generate the pose error compensation amount.

[0017] Optionally, the step of processing the height deviation signal and the tilt deviation signal based on the height channel gain coefficient and the tilt channel gain coefficient using a closed-loop control algorithm to generate a pose error compensation amount includes:

[0018] Using a closed-loop control algorithm, the height deviation signal is time-varying weighted by the height channel gain coefficient, and the tilt deviation signal is time-varying weighted by the tilt channel gain coefficient, to generate a weighted height deviation signal and a weighted tilt deviation signal.

[0019] Dead zone crossing processing is performed on the weighted height deviation signal and the weighted tilt angle deviation signal to generate height compensation component and tilt angle compensation component;

[0020] The height compensation component and the tilt compensation component are combined to form the pose error compensation amount.

[0021] Optionally, the step of performing dead-zone crossing processing on the weighted height deviation signal and the weighted tilt angle deviation signal to generate height compensation components and tilt angle compensation components includes:

[0022] Mechanical vibration data is collected by an inertial measurement unit installed on the multi-degree-of-freedom actuator, and the peak vibration energy is calculated based on the mechanical vibration data.

[0023] When the peak vibration energy exceeds the preset vibration threshold, the weighted height deviation signal and the weighted tilt angle deviation signal are attenuated using a preset vibration suppression coefficient to obtain the height correction value and the tilt angle correction value.

[0024] Based on the motion state of the multi-degree-of-freedom actuator, dynamically calculate the height dead zone threshold and tilt angle dead zone threshold;

[0025] In the nonlinear dead zone compensation process, if the height correction value is greater than the height dead zone threshold, a height compensation component is generated based on the height correction value using a preset height compensation formula. If the tilt angle correction value is greater than the tilt angle dead zone threshold, a tilt angle compensation component is generated based on the tilt angle correction value using a preset tilt angle compensation formula.

[0026] Optionally, the dynamic mapping model based on preset density, terrain, and pose maps the density distribution data and the terrain feature data to the target pose parameters of the controlled object, including:

[0027] Spatial density focusing operation is performed on the density distribution data to generate a set of coordinates for the density core region; curvature decomposition operation is performed on the terrain feature data to generate a terrain gradient tensor.

[0028] Based on the density core region coordinate set and the terrain gradient tensor, the initial height value is generated using the height mapping function in the dynamic mapping model, which is used to characterize the relationship between density, terrain and height.

[0029] Based on the density core region coordinate set and the terrain gradient tensor, the initial value of the dip angle is generated using the dip angle mapping function in the dynamic mapping model, which is used to characterize the relationship between density, terrain and dip angle.

[0030] Based on the compensation coefficient corresponding to the growth stage of the target object, a compensation operation is performed on the initial height value to generate a height setting value. Based on the influence of the ambient wind speed on the target object, an adaptive compensation operation is performed on the initial tilt angle value to generate a tilt angle setting value.

[0031] The target pose parameters are formed by combining the height setting value and the tilt angle setting value.

[0032] Optionally, the step of performing a spatial density focusing operation on the density distribution data to generate a set of coordinates for the density core region includes:

[0033] Perform a longitudinal density projection operation on the density distribution data to generate a longitudinal density distribution curve;

[0034] Perform bimodal detection on the longitudinal density distribution curve to identify the first density core region and the second density core region;

[0035] A radial density scan operation is performed within the first density core region and the second density core region to generate a first core point coordinate set and a second core point coordinate set.

[0036] The first core point coordinate set and the second core point coordinate set are merged to form the density core region coordinate set.

[0037] Optionally, generating the pose control signal based on the pose error compensation amount includes:

[0038] The pose error compensation amount is input into the servo signal converter to perform a control amount allocation operation and generate a first height control amount and a first tilt angle control amount.

[0039] Perform anti-saturation constraint operation on the first height control value and the first tilt angle control value to generate the second height control value and the second tilt angle control value.

[0040] Based on the second height control quantity, a corresponding servo valve control signal is generated; based on the second tilt angle control quantity, a pulse width modulation signal corresponding to the tilt angle is generated using a preset electromechanical coupling equation.

[0041] The servo valve control signal and the pulse width modulation signal are combined to form a pose control signal.

[0042] Secondly, this application provides an operation control system for a crop harvester, comprising:

[0043] The acquisition module is used to acquire density distribution data of the target object and terrain feature data of the area where the target object is located;

[0044] The mapping module is used to map the density distribution data and the terrain feature data into target pose parameters of the controlled object based on a preset dynamic mapping model of density, terrain and pose. The target pose parameters include: height setting value and tilt angle setting value. The controlled object is a multi-degree-of-freedom actuator of a crop harvester.

[0045] The input module is used to obtain the actual pose parameters of the controlled object, input the actual pose parameters and the target pose into the adaptive controller, and dynamically calculate the pose error compensation amount through the closed-loop control algorithm.

[0046] The drive module is used to generate a pose control signal based on the pose error compensation amount, and drive the pose adjustment structure to perform pose calibration on the controlled object, so as to adjust the working height and tilt angle of the end effector installed at the front end of the multi-degree-of-freedom actuator.

[0047] Thirdly, this application provides a computing device, including a processing component and a storage component; the storage component stores one or more computer instructions; the one or more computer instructions are invoked and executed by the processing component to implement an operation control method for a crop harvester as described in any of the first aspects.

[0048] Fourthly, this application provides a computer storage medium storing a computer program, which, when executed by a computer, implements an operation control method for a crop harvester as described in any of the first aspects.

[0049] This application provides a method for controlling the operation of a crop harvester. The method includes: acquiring density distribution data of a target object and terrain feature data of the area where the target object is located; mapping the density distribution data and terrain feature data to target pose parameters of the controlled object based on a preset dynamic mapping model of density, terrain, and pose, wherein the target pose parameters include a height setpoint and an inclination angle setpoint, and the controlled object is a multi-degree-of-freedom actuator of the crop harvester; acquiring the actual pose parameters of the controlled object, inputting the actual pose parameters and the target pose into an adaptive controller, and dynamically calculating the pose error compensation amount through a closed-loop control algorithm; generating a pose control signal based on the pose error compensation amount, driving a pose adjustment structure to perform pose calibration on the controlled object, so as to adjust the working height and inclination angle of the end effector installed at the front end of the multi-degree-of-freedom actuator.

[0050] This application provides an objective basis for precise control by acquiring crop density distribution and terrain feature data of the target area. Then, using a pre-set dynamic mapping model, the perceived density and terrain information are intelligently converted into the height and tilt angle settings required by the multi-degree-of-freedom actuator, achieving intelligent decision-making on the target posture based on crop status and terrain. The actual posture of the actuator is acquired in real time and compared with the target value, then input into an adaptive controller for closed-loop dynamic calculation, generating precise posture error compensation, effectively overcoming internal and external disturbances. Finally, based on the compensation, a control signal is generated to drive the posture adjustment structure, achieving real-time, dynamic calibration of the multi-degree-of-freedom actuator. This precisely controls the working height and tilt angle of the end effector, ensuring it dynamically matches changing crop density and terrain undulations, guaranteeing the consistency and high quality of harvesting operations.

[0051] Furthermore, this application first separates the actual height and tilt angle values ​​from the actual pose parameters of the actuator; then, it compares the set height value with the actual height value to generate a height deviation signal, and compares the set tilt angle value with the actual tilt angle value to generate a tilt angle deviation signal; then, using a multi-parameter coupling compensation module, combined with the actuator's dynamic transfer function, it performs disturbance estimation on these two deviation signals to obtain pose disturbance components containing height and tilt angle disturbance components; subsequently, it compares the estimated height disturbance component and tilt angle disturbance component with preset stability boundary conditions, and dynamically adjusts the height channel gain coefficient and tilt angle channel gain coefficient accordingly; finally, through a closed-loop control algorithm, it uses the adjusted gain coefficient to perform time-varying weight scaling on the deviation signals of each channel to generate a weighted deviation signal, and then performs dead-zone crossing processing to eliminate the influence of nonlinear intervals, ultimately generating height compensation components and tilt angle compensation components, which are combined to form the final pose error compensation amount.

[0052] Furthermore, by estimating the coupled disturbance components of the height and tilt channels in real time and dynamically and adaptively adjusting the control gain coefficients of each channel according to the disturbance intensity, the anti-disturbance capability and robustness of the control system are improved. The time-varying weight scaling strategy enables the controller to intelligently allocate control force according to the current disturbance state, more effectively suppressing coupled interference. The introduction of dead-zone crossing processing effectively eliminates the oscillations or sluggish response caused by the actuator in the nonlinear range near zero, improving the smoothness and accuracy of the system response. Finally, the generated pose error compensation can more accurately and collaboratively compensate for height and tilt errors, ensuring that the end effector pose can quickly and stably track the target setpoint under complex disturbances.

[0053] These or other aspects of this application will become more apparent in the following description of the embodiments. Attached Figure Description

[0054] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0055] Figure 1 A flowchart illustrating an operation control method for a crop harvester provided in this application embodiment;

[0056] Figure 2 This is a schematic diagram of the operation control system of a crop harvester provided in an embodiment of this application;

[0057] Figure 3 This is a schematic diagram of the structure of a computing device provided in an embodiment of this application. Detailed Implementation

[0058] To enable those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings.

[0059] In some of the processes described in the specification, claims, and accompanying drawings of this application, multiple operations appearing in a specific order are included. However, it should be clearly understood that these operations may not be executed in the order they appear herein, or may be executed in parallel. The operation numbers, such as 11, 12, etc., are merely used to distinguish different operations and do not themselves represent any execution order. Furthermore, these processes may include more or fewer operations, and these operations may be executed sequentially or in parallel. It should be noted that the descriptions such as "first," "second," etc., in this document are used to distinguish different messages, devices, modules, etc., and do not represent a sequential order, nor do they limit "first" and "second" to different types.

[0060] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0061] To address the problems of inaccurate control of harvesting height and tilt angle in existing technologies, and the inability to dynamically respond to changes in the field, this application provides an operation control method for a crop harvester. This method employs the following concept: First, it synchronously acquires objective environmental information of the target area through sensors. Second, it innovatively constructs and pre-sets a dynamic mapping model, which intelligently integrates density and terrain as key inputs to directly derive the optimal target pose parameters required by the actuator, achieving a transformation from perception to decision-making. Then, it introduces an adaptive controller to form a closed loop, dynamically comparing the actual pose of the actuator with the target pose by real-time monitoring, and accurately calculating the error compensation amount using a closed-loop control algorithm. Finally, it converts the compensation amount into a control signal, driving the pose adjustment structure on the actuator to perform real-time, proactive calibration, thereby dynamically and collaboratively adjusting the height and tilt angle of the end effector, ensuring that its working posture always adaptively matches the density and terrain conditions of the current working area, guaranteeing consistent and high-quality harvesting.

[0062] Figure 1 A flowchart illustrating an operation control method for a crop harvester provided in this application embodiment is shown below. Figure 1 As shown, the method includes:

[0063] S11. Obtain the density distribution data of the target object and the terrain feature data of the area where the target object is located.

[0064] The target object can refer to the crop plant population within the current work area, such as corn, wheat, and leeks. It's important to note that the same crop in different work areas may have different surface growing environments. Density distribution data can refer to a two-dimensional matrix reflecting the density of crop plants per unit area, acquired through optical sensors. Topographic feature data can refer to a set of topographic parameters, including surface elevation, slope, and curvature, generated through 3D scanning.

[0065] In this embodiment, a multispectral sensor installed at the front end of the crop harvester first scans the target object, i.e., the crop plant group, in real time to obtain density distribution data reflecting the density of the plants; simultaneously, a lidar is used to perform a three-dimensional terrain scan of the area where the target object is located, extracting surface undulation features to form terrain feature data. These two types of data can be synchronously transmitted to the central processing unit via the vehicle bus.

[0066] S12. Based on the preset dynamic mapping model of density, terrain and pose, the density distribution data and terrain feature data are mapped to the target pose parameters of the controlled object. The target pose parameters include: height setting value and tilt angle setting value. The controlled object is the multi-degree-of-freedom actuator of a crop harvester.

[0067] The preset dynamic mapping model between density, terrain, and pose can refer to a mathematical model based on machine learning that maps crop density and terrain features to the target pose of the actuator. The controlled object can refer to the multi-degree-of-freedom actuator of a crop harvester that requires dynamic adjustment. The multi-degree-of-freedom actuator can refer to a mechanical device with lifting and pitching functions, consisting of hydraulic cylinders and a slewing mechanism. The target pose parameters can refer to the set of spatial pose target values ​​that the actuator needs to achieve. The height setting value can refer to the target vertical distance reference of the end effector relative to the ground surface. The tilt angle setting value can refer to the target angle reference between the end effector's cutting edge plane and the horizontal plane.

[0068] In this embodiment, the central processing unit invokes a preset dynamic mapping model of density terrain and pose, which is a neural network trained through field experiments. First, density distribution data is input into the model's height mapping channel, and combined with the elevation change rate from the terrain feature data, a weighted fusion calculation is performed to generate a height setpoint. Simultaneously, density distribution data is input into the tilt angle mapping channel, and combined with the slope angle from the terrain feature data, a nonlinear transformation is performed to generate a tilt angle setpoint. Finally, the target pose parameters, including the height and tilt angle setpoints, are output, specifying the target spatial pose of the controlled object, i.e., the multi-degree-of-freedom actuator.

[0069] S13. Obtain the actual pose parameters of the controlled object, input the actual pose parameters and the target pose into the adaptive controller, and dynamically calculate the pose error compensation amount through the closed-loop control algorithm.

[0070] The actual pose parameters refer to the true values ​​of the actuator's spatial pose measured in real time by sensors. An adaptive controller is an intelligent control module that can automatically adjust control parameters based on the system state. A closed-loop control algorithm refers to a control logic implementation method that continuously corrects the output through feedback signals. The pose error compensation amount refers to the amount of spatial pose correction required to eliminate the difference between the actual pose and the target pose.

[0071] In this embodiment, an encoder installed at the joint of a multi-degree-of-freedom actuator collects actual pose parameters in real time, including the actual height value converted from the hydraulic cylinder extension / retraction amount and the actual tilt angle value measured by an attitude sensor. The actual pose parameters and target pose parameters are input to an adaptive controller. Inside the controller, a first difference between the set height value and the actual height value is calculated to form a height deviation signal, and a second difference between the set tilt angle value and the actual tilt angle value is calculated to form a tilt angle deviation signal. Subsequently, a fuzzy closed-loop control algorithm is used to dynamically adjust the gain and compensate for coupling interference of the two types of deviation signals, ultimately outputting the pose error compensation amount.

[0072] S14. Generate a pose control signal based on the pose error compensation amount, drive the pose adjustment structure to perform pose calibration on the controlled object, so as to adjust the working height and tilt angle of the end effector installed at the front end of the multi-degree-of-freedom actuator.

[0073] In this context, "position control signal" refers to the set of electrical control commands that drive the actuator. "Position adjustment structure" refers to a position execution device that includes a hydraulic valve assembly and an electric actuator. "End effector," or cutter, refers to the cutting tool assembly installed at the front end of the actuator to perform the cutting function. "Working height" can refer to the vertical working distance between the cutting edge and the root of the crop. "Inclination angle" can refer to the working angle formed between the cutting plane and the horizontal plane.

[0074] In this embodiment, the pose error compensation amount is input to the signal conversion module, which first decomposes it into control components for the height axis and tilt axis. Then, a pose control signal for driving the pose adjustment structure is generated using pulse width modulation technology. This signal includes a current signal controlling the opening of the hydraulic valve and a pulse width modulation signal controlling the electric actuator. Finally, the hydraulic actuator and electric actuator in the pose adjustment structure are driven to work together to perform spatial pose calibration on the multi-degree-of-freedom actuator, so that the working height and tilt angle of the end effector installed at its front end are consistent with the target set value.

[0075] Here is a specific example: When a crop harvester moves to area A, it first scans the crop plants within a 3-meter radius ahead using an onboard spectral imager to generate a density distribution heatmap; simultaneously, a lidar collects the surface elevation data for this area. Next, the central processing unit inputs the density heatmap and elevation data into a pre-trained pose mapping model, outputting a height setting of 4 cm and an inclination setting of 8 degrees. The height setting is based on the typical stubble height of mature crops (2 cm) plus the 2 cm gap required to protect the stem base. The inclination setting is calculated based on the current terrain slope of approximately 10 degrees; that is, it is the product of the slope angle and a coefficient of 0.8. Subsequently, the joint encoder detects that the actual height of the actuator is 3 cm and the inclination angle is 7 degrees. The controller calculates a height deviation of 1 cm and an inclination angle deviation of 1 degree, which are then used by a fuzzy proportional-integral-derivative (PID) controller to generate compensation values. Finally, the compensation amount is converted into a hydraulic valve opening signal and a push rod pulse width modulation (PWM) signal, which drives the hydraulic cylinder to rise by 1 cm while the electric push rod adjusts by 1 degree, so that the cutter accurately maintains a height of 4 cm and an inclination angle of 8 degrees when entering the working area.

[0076] By executing S11~S14, this embodiment of the application intelligently generates the target pose parameters of the actuator by synchronously sensing crop density and terrain features; it uses adaptive closed-loop control to compensate for pose errors in real time, and dynamically drives the adjustment mechanism to coordinately adjust the height and tilt angle of the end effector, so that the cutting posture always adapts to the changes in field density distribution and terrain undulations, effectively improving the crop harvesting integrity rate and reducing root and stem damage.

[0077] In one possible embodiment, S13, inputting the actual pose parameters and the target pose into the adaptive controller, and dynamically calculating the pose error compensation amount through a closed-loop control algorithm, includes:

[0078] Step 131: In the adaptive controller, obtain the actual height value and the actual tilt angle value from the actual pose parameters.

[0079] The actual height value can refer to the true value of the vertical position of the cutter as measured in real time by the displacement sensor of the actuator. The actual tilt angle value can refer to the real-time measured value of the angle between the end effector and the horizontal plane obtained by the inertial measurement unit.

[0080] In this embodiment, the adaptive controller first receives the actual pose parameter data packet from the multi-degree-of-freedom actuator, and then extracts the actual height value measured by the joint encoder from it through the data parsing module, while separating the actual tilt angle value collected by the attitude sensor.

[0081] Step 132: Track the set height value and the actual height value using a setpoint to generate a height deviation signal. Track the set tilt angle value and the actual tilt angle value using a setpoint to generate a tilt angle deviation signal.

[0082] Among these, setpoint tracking refers to the control process that continuously compares the target setpoint with the actual measured value to generate an error signal. The height deviation signal refers to the vertical position error obtained by subtracting the actual height value from the height setpoint. The tilt angle deviation signal refers to the angular error obtained by subtracting the actual tilt angle value from the tilt angle setpoint.

[0083] In this embodiment, the height setting value is first input to a setpoint tracker and compared in real time with the extracted actual height value. A height deviation signal is generated by subtraction. Simultaneously, the tilt angle setting value is input to another setpoint tracker and subtracted from the actual tilt angle value to generate a tilt angle deviation signal.

[0084] Step 133: Through the multi-parameter coupling compensation module of the adaptive controller, combined with the dynamic transfer function of the multi-degree-of-freedom actuator, disturbance estimation is performed on the height deviation signal and tilt angle deviation signal to obtain the pose disturbance component containing the height disturbance component and the tilt angle disturbance component.

[0085] The multi-parameter coupling compensation module can refer to a dedicated algorithm unit used to estimate dual-channel disturbances and decouple control. The dynamic transfer function can refer to a differential equation model describing the input-output relationship of the actuator's mechanical system. This dynamic transfer function explicitly characterizes the inertial and elastic coupling mechanism of the height and tilt channels through an asymmetric coupling matrix. Based on acceleration residual calculation technology, the multi-parameter coupling compensation module uses deviation signals to separate the pose disturbance components in real time, providing accurate disturbance estimates for subsequent adaptive gain adjustment. Disturbance estimation can refer to the calculation process of separating system disturbance factors through an observer algorithm. The height disturbance component can refer to the disturbance estimate affected by mechanical vibration and load changes in the vertical direction. The tilt disturbance component can refer to the disturbance estimate affected by ground impact and inertial forces in the rotational direction. The pose disturbance component includes a composite estimate of both height and tilt channel disturbances. The formula for the dynamic transfer function is:

[0086] ,in, Input force for height channel control, Input torque is used to control the tilt angle channel. For the equivalent mass in the height direction, The inertial coupling coefficient is the inclination angle and height. The inertial coupling coefficient for height and tilt angle. The moment of inertia is the rotational inertia in the tilt direction. For the linear acceleration of the end effector, For the end effector angular acceleration, The damping coefficient is in the height direction. The damping coupling coefficient for height and tilt angle. The damping coupling coefficient is the inclination angle and height. The damping coefficient is in the tilt direction. For the rate of change of height, For the angular velocity of the tilt angle change, Here is the stiffness coefficient in the height direction. This is the stiffness coefficient in the tilt direction. This represents the actual height of the end effector. This represents the actual tilt angle of the end effector. External disturbances to the high-altitude channel This refers to external disturbances in the tilt channel.

[0087] In this embodiment, the height deviation signal and tilt angle deviation signal are first received through the multi-parameter coupling compensation module. Then, the dynamic transfer function pre-stored by the multi-degree-of-freedom actuator is called, and the extended Kalman filter algorithm is used to perform joint disturbance estimation on the dual-channel deviation signal. Finally, the pose disturbance component containing the height disturbance component and the tilt angle disturbance component is output.

[0088] Step 134: Compare the height disturbance component and the tilt disturbance component with the corresponding preset stability boundary conditions to obtain the dynamically adjusted height channel gain coefficient and tilt channel gain coefficient.

[0089] Among these, stability boundary conditions can refer to the set of maximum permissible disturbance thresholds that ensure the stability of the control system. The height channel gain coefficient can refer to the scaling factor that dynamically adjusts the sensitivity of the vertical control loop based on the disturbance intensity. The tilt channel gain coefficient can refer to the scaling factor that dynamically adjusts the sensitivity of the rotary control loop based on the disturbance intensity.

[0090] In this embodiment, a preset stability boundary condition database is first read. Then, the amplitude of the height disturbance component is compared with that of the height channel stability boundary condition, and the height channel gain coefficient is dynamically calculated based on the over-limit ratio. Simultaneously, the tilt disturbance component is compared with that of the tilt channel stability boundary condition, and the tilt channel gain coefficient is adaptively adjusted according to the disturbance intensity.

[0091] Step 135: Using a closed-loop control algorithm, based on the height channel gain coefficient and the tilt channel gain coefficient, process the height deviation signal and the tilt deviation signal to generate the pose error compensation amount.

[0092] In this embodiment, a proportional-integral module of a closed-loop control algorithm is first used to weight and amplify the height deviation signal using the height channel gain coefficient, and simultaneously, the tilt deviation signal is weighted and amplified using the tilt channel gain coefficient. Then, coupling interference compensation calculations are performed on the weighted dual-channel signals, and finally, the signals are fused to generate the pose error compensation amount.

[0093] Here is a specific example: When the harvester is operating on a slope, the controller first parses the actual height value of 4.2 cm and the actual tilt angle value of 8 degrees from the posture parameters. Next, it compares the height setpoint of 5.0 cm with the actual height value to generate a height deviation signal of 0.8 cm, and simultaneously compares the tilt angle setpoint of 12 degrees with the actual tilt angle value to generate a tilt angle deviation signal of 4 degrees. Then, the multi-parameter coupling compensation module, combined with the hydraulic system transfer function, estimates the height disturbance component of 0.3 cm and the tilt angle disturbance component of 1.5 degrees. Subsequently, it compares the height disturbance component with a preset threshold of 0.5 to generate a height gain coefficient of 1.3 times, and compares the tilt angle disturbance component with a preset threshold of 1.0 degrees to generate a tilt angle gain coefficient of 1.8 times. Finally, it amplifies the deviation signals using the gain coefficients, and through PID calculation, outputs a drive to extend the hydraulic cylinder by 0.7 cm while simultaneously adjusting the slewing mechanism by a compensation amount of 3.2 degrees.

[0094] By executing steps 131 to 135, this embodiment of the application estimates the dual-channel disturbance components in real time through decoupling, and dynamically optimizes the control gain based on the stability boundary, effectively suppressing coupling interference during the motion of the actuator, and improving the pose tracking accuracy and system robustness.

[0095] In one possible embodiment, step 135, processing the height deviation signal and tilt deviation signal based on the height channel gain coefficient and tilt angle channel gain coefficient using a closed-loop control algorithm, and generating a pose error compensation amount, includes:

[0096] Step a1: Using a closed-loop control algorithm, the height deviation signal is time-varyingly weighted and scaled using the height channel gain coefficient, and the tilt deviation signal is time-varyingly weighted and scaled using the tilt channel gain coefficient, to generate a weighted height deviation signal and a weighted tilt deviation signal.

[0097] Time-varying weighted scaling refers to the process of dynamically adjusting the amplitude of the original signal based on the gain coefficient that changes in real time according to the system state. The weighted height deviation signal refers to the vertical position error after scaling by the height channel gain coefficient. The weighted tilt deviation signal refers to the angular error after scaling by the tilt channel gain coefficient.

[0098] In this embodiment, the closed-loop control algorithm first receives the height channel gain coefficient and tilt channel gain coefficient calculated by the previous stage. Then, the height channel gain coefficient is used to dynamically amplify or attenuate the original height deviation signal in real time; specifically, the height deviation signal is multiplied by this time-varying coefficient to generate a weighted height deviation signal. Simultaneously, the tilt channel gain coefficient is used to dynamically scale the original tilt deviation signal in real time; the tilt deviation signal is then multiplied by this time-varying coefficient to generate a weighted tilt deviation signal.

[0099] Step a2: Perform dead zone crossing processing on the weighted height deviation signal and the weighted tilt angle deviation signal to generate height compensation component and tilt angle compensation component.

[0100] Dead zone crossing processing refers to the algorithmic process of filtering out minute jitter signals and converting them into nonlinear signals by using a preset threshold. Height compensation component refers to the effective vertical control correction amount after eliminating the effects of mechanical static friction. Tilt angle compensation component refers to the effective angular control correction amount for overcoming the resistance of the slewing mechanism.

[0101] In this embodiment, a preset dead zone threshold parameter is first read, and then the weighted height deviation signal is input to a nonlinear processing unit. When the absolute value of the signal is less than the height dead zone threshold, a zero value is output; when it exceeds the threshold, a height compensation component is generated according to a preset function. Simultaneously, the weighted tilt angle deviation signal is input to another nonlinear processing unit; when it is less than the tilt angle dead zone threshold, it is reset to zero; when it exceeds the threshold, a tilt angle compensation component is generated based on the compensation curve.

[0102] Step a3: Combine the height compensation component and the tilt compensation component to form the pose error compensation amount.

[0103] In this embodiment, the height compensation component and tilt compensation component are first input into a multi-channel synthesizer. Then, vector superposition calculation is performed according to the spatial pose transformation relationship. Specifically, the height compensation component is used as a vertical direction correction, and the tilt compensation component is used as a rotation direction correction for orthogonal synthesis. The final output is a pose error compensation data packet containing two degrees of freedom compensation quantities.

[0104] Here's a specific example: When a harvester operates on bumpy roads, the controller first applies a 1.2x height gain factor to scale a 1.5cm height deviation into a 1.8cm weighted height deviation signal, and simultaneously applies a 1.5x tilt gain factor to scale a 2-degree tilt deviation into a 3.0-degree weighted tilt deviation signal. Next, if the height signal exceeds a 0.3cm dead zone threshold, a 1.5cm height compensation component is generated using a linear transformation formula; similarly, if the tilt signal exceeds a 0.5-degree threshold, a 2.2-degree tilt compensation component is generated. Finally, the vertical compensation and rotational compensation are combined using a spatial vector to output a 1.5cm extension of the hydraulic cylinder and a 2.2-degree adjustment of the slewing mechanism's pose error compensation.

[0105] By executing steps a1 to a3, the embodiments of this application optimize the error response sensitivity through dynamic gain scaling, eliminate nonlinear interference by combining dead zone processing, generate accurate and effective two-degree-of-freedom compensation quantities, and improve the position control accuracy and stability of the actuator.

[0106] In one possible embodiment, step a2, performing dead-zone crossing processing on the weighted height deviation signal and the weighted tilt angle deviation signal to generate height compensation components and tilt angle compensation components, includes:

[0107] Step a21: Collect mechanical vibration data based on the inertial measurement unit installed on the multi-degree-of-freedom actuator, and calculate the peak vibration energy based on the mechanical vibration data.

[0108] In this context, the inertial measurement unit (IMU) can refer to a sensor module integrating a three-axis gyroscope and an accelerometer, used to collect the spatial motion parameters of the actuator. Mechanical vibration data can refer to a time-series dataset of accelerations reflecting structural vibrations during equipment operation. Vibration energy peak value can refer to the maximum energy value corresponding to the dominant vibration frequency extracted through spectral analysis.

[0109] In this embodiment, triaxial acceleration data is first collected in real time by an inertial measurement unit installed at a key node of the multi-degree-of-freedom actuator, forming a mechanical vibration data stream. Then, a fast Fourier transform is performed on the mechanical vibration data to extract the vibration energy spectrum of the main frequency bands, and the peak vibration energy is calculated by integration.

[0110] Step a22: When the peak vibration energy exceeds the preset vibration threshold, the weighted height deviation signal and the weighted tilt angle deviation signal are attenuated using the preset vibration suppression coefficient to obtain the height correction value and the tilt angle correction value.

[0111] Among them, the preset vibration threshold can refer to the critical value of safe vibration intensity set according to the vibration resistance capability of the mechanism. The preset vibration suppression coefficient can refer to the vibration interference attenuation ratio factor calibrated according to experiments. Attenuation processing can refer to the interference suppression process that reduces the signal amplitude using multiplication operations. The height correction value can refer to the effective vertical control quantity after vibration suppression processing. The tilt angle correction value can refer to the effective rotation control quantity after vibration suppression processing.

[0112] In this embodiment, the peak vibration energy is first compared with a preset vibration threshold in real time. When the peak vibration energy exceeds the threshold, a preset vibration suppression coefficient database is invoked, and the weighted height deviation signal is attenuated by multiplication to obtain a height correction value. At the same time, the weighted tilt angle deviation signal is attenuated proportionally to obtain a tilt angle correction value.

[0113] Step a23: Calculate the height dead zone threshold and tilt dead zone threshold dynamically based on the motion state of the multi-degree-of-freedom actuator.

[0114] Here, motion state can refer to dynamic operating parameters including the real-time speed and acceleration of the actuator. Height dead zone threshold can refer to the vertical control sensitivity threshold dynamically adjusted based on motion speed. Tilt angle dead zone threshold can refer to the rotational control sensitivity threshold dynamically adjusted based on motion acceleration.

[0115] In this embodiment, the current motion speed and acceleration parameters are first obtained through the joint encoder of the actuator, and the motion state level is defined. Then, the height dead zone threshold is calculated using linear interpolation based on the speed parameters, and the tilt angle dead zone threshold is calculated using a nonlinear mapping relationship based on the acceleration parameters.

[0116] Step a24: In the nonlinear dead zone compensation process, if the height correction value is greater than the height dead zone threshold, then the height compensation component is generated based on the height correction value using the preset height compensation formula. If the tilt correction value is greater than the tilt dead zone threshold, then the tilt compensation component is generated based on the tilt correction value using the preset tilt compensation formula.

[0117] The nonlinear dead zone compensation process refers to the algorithm flow that filters out minute errors and transforms signals in the nonlinear region. The preset height compensation formula refers to the vertical displacement transformation function that eliminates the influence of static friction in the hydraulic system. The preset tilt angle compensation formula refers to the angle transformation function that overcomes the backlash effect of the rotary mechanism.

[0118] In this embodiment, during the nonlinear dead zone compensation process, the absolute value of the height correction value is first compared with the dynamic height dead zone threshold. When the height correction value exceeds the threshold, it is input into a preset height compensation formula for nonlinear transformation to generate a height compensation component. Simultaneously, the tilt correction value is compared with the dynamic tilt dead zone threshold; if it exceeds the threshold, a tilt compensation component is generated using a preset tilt compensation formula.

[0119] Here is a specific example: When a harvester operates on soft, muddy ground after rain, the inertial measurement unit first detects a vibration energy peak of 1.2g exceeding a preset threshold of 1.0g. Next, a vibration suppression coefficient of 0.8 is used to attenuate the 1.5cm weighted height deviation to a 1.2cm height correction value, and the 1.8-degree weighted tilt angle deviation to a 1.4° tilt angle correction value. Then, a 0.3cm height dead zone threshold is calculated based on the current movement speed of 0.6m / s, and a 0.5-degree tilt angle dead zone threshold is calculated based on an acceleration of 0.4m / s². Finally, when the height correction value exceeds the threshold, a 0.9cm height compensation component is generated; when the tilt angle correction value exceeds the threshold, a 1.1-degree tilt angle compensation component is generated. By executing steps a21 to a24, this embodiment of the application dynamically suppresses mechanical interference through real-time vibration monitoring and adaptively adjusts the dead zone threshold based on the motion state, effectively eliminating the influence of nonlinear links and improving the vibration resistance and dynamic response accuracy of the posture control system.

[0120] In one possible embodiment, S12, based on a preset dynamic mapping model of density, terrain, and pose, maps density distribution data and terrain feature data into target pose parameters of the controlled object, including:

[0121] Step 121: Perform spatial density focusing operation on the density distribution data to generate a set of coordinates for the density core region; perform curvature decomposition operation on the terrain feature data to generate a terrain gradient tensor.

[0122] The spatial density focusing operation refers to the data processing procedure of extracting the center coordinates of high-density crop regions using a clustering algorithm. The density core region coordinate set refers to the set of latitude and longitude coordinates representing the location of the center point of the densely growing crop region. In this embodiment, the density core region coordinate set can be represented as:

[0123] The union of {(2.3,4.1),(2.5,4.3),(2.7,4.0),(3.0,4.2),(2.8,4.5)} and {(7.1,3.8),(7.3,3.9),(7.5,4.0),(7.6,4.2),(7.4,4.3)}. Curvature decomposition refers to a method of calculating the principal curvature directions by performing differential geometric analysis on a terrain surface. The terrain gradient tensor refers to a second-order matrix data structure describing the rate and direction of change of surface slope.

[0124] In this embodiment, a spatial density focusing operation is first performed on the density distribution data, and the center points of high-density areas are identified using a Gaussian mixture model clustering algorithm to generate a set of coordinates for the density core area. Simultaneously, a curvature decomposition operation is performed on the terrain feature data, and the eigenvectors of the surface curvature tensor are calculated using differential geometry methods to form a terrain gradient tensor describing the direction and intensity of slope changes.

[0125] Step 122: Based on the density core region coordinate set and the terrain gradient tensor, generate initial height values ​​using the height mapping function in the dynamic mapping model, which is used to characterize the relationship between density, terrain, and height.

[0126] The height mapping function refers to a mathematical model that maps crop density characteristics to topographic parameters as a reference height. The initial height value refers to a preliminary height setting reference that does not consider growth stage compensation. The formula for the height mapping function is:

[0127] ,in, The initial height value, Density sensitivity coefficient For the number of core points, The coordinates of the density core region are... As the reference point coordinates, This is the density attenuation coefficient. For terrain sensitivity coefficient, For the terrain gradient tensor, For the largest eigenvector, This is the reference height.

[0128] In this embodiment, the density core region coordinate set and the terrain gradient tensor are first input into the dynamic mapping model. Then, the height mapping function is called, which multiplies the longitudinal distribution density of the core region coordinates by the vertical component coefficient of the terrain gradient tensor, and then superimposes the reference height parameter to finally output the initial height value.

[0129] Step 123: Based on the density core region coordinate set and the terrain gradient tensor, generate the initial value of the dip angle using the dip angle mapping function in the dynamic mapping model, which is used to characterize the relationship between density, terrain and dip angle.

[0130] The tilt mapping function refers to a mathematical model that maps crop distribution characteristics to topographic parameters as a reference angle; the initial tilt value refers to the preliminary tilt setting reference without considering wind speed compensation. The formula for the tilt mapping function is:

[0131] ;

[0132] in, The initial target inclination angle, Density direction factor For the number of core points, For core point density weight, As the azimuth of the core point, For terrain sensitivity coefficient, For topographic curvature, The magnitude of the maximum curvature vector. The minimum curvature vector magnitude, The reference tilt angle is used.

[0133] In this embodiment, the same density core region coordinate set and terrain gradient tensor are first input into the dynamic mapping model. Then, the tilt mapping function is called, which divides the lateral dispersion of the core region coordinates by the horizontal component magnitude of the terrain gradient tensor, multiplies it by the angle transformation factor, and finally outputs the initial tilt angle value.

[0134] Step 124: Based on the compensation coefficient corresponding to the growth stage of the target object, perform a compensation operation on the initial height value to generate a height setting value. Based on the influence of the ambient wind speed on the target object, perform an adaptive compensation operation on the initial tilt angle value to generate a tilt angle setting value.

[0135] The target object's growth stage can refer to the growth status of crops, such as seedling stage or maturity stage, determined by visual characteristics. The compensation coefficient can refer to the adjustment factor based on the height setting value calibrated according to the growth stage. The compensation operation can refer to the arithmetic correction process of multiplying the initial height value by the compensation coefficient. The ambient wind speed can refer to the real-time wind speed value of the work area obtained through meteorological sensors. The adaptive compensation operation can refer to the calculation process of nonlinearly correcting the initial tilt angle value based on the wind speed.

[0136] In this embodiment, the same density core region coordinate set and terrain gradient tensor are first input into the dynamic mapping model. Then, the tilt mapping function is called, which divides the lateral dispersion of the core region coordinates by the horizontal component magnitude of the terrain gradient tensor, multiplies it by the angle transformation factor, and finally outputs the initial tilt angle value.

[0137] Step 125: Combine the height setting value and tilt angle setting value to form the target pose parameters.

[0138] In this embodiment, the height and tilt angle settings are first input into a data encapsulation module. Then, structured encoding is performed according to the spatial pose parameter specification, ultimately outputting a complete target pose parameter data package containing the two-degree-of-freedom settings.

[0139] Here's a specific example: When the harvester enters area B, firstly, cluster analysis is performed on the density distribution map to generate a coordinate set containing 8 core points. Simultaneously, curvature decomposition of the terrain data yields a gradient tensor describing an 8-degree slope. Next, the coordinate set and tensor are input into a height mapping function, which outputs an initial height value of 4.2 cm, and into a tilt angle mapping function, which outputs an initial tilt angle value of 10 degrees. Subsequently, based on image recognition detecting the seedling stage, a compensation coefficient of 1.05 is used to increase the height value by 4.4 cm, while the tilt angle value is compensated to 12 degrees based on a wind speed of 3 meters per second. Finally, the 4.4 cm height setting and the 12-degree tilt angle setting are combined to form the target pose parameters.

[0140] By executing steps 121 to 125, this embodiment of the application generates an initial pose value by intelligently fusing the core features of crop density with terrain gradient parameters, and performs dynamic compensation by combining the growth stage and wind speed influence, so that the target pose parameters are more accurately adapted to the actual working environment.

[0141] In one possible embodiment, step 121, performing a spatial density focusing operation on the density distribution data to generate a density core region coordinate set, includes:

[0142] Step b1: Perform a longitudinal density projection operation on the density distribution data to generate a longitudinal density distribution curve.

[0143] The longitudinal density projection operation refers to a data processing method that generates a one-dimensional distribution curve by performing column integration on a two-dimensional density map along the harvester's travel direction. The longitudinal density distribution curve can be defined as a continuous function of peaks and troughs reflecting the trend of crop density variation along the travel direction.

[0144] In this embodiment, a vertical density projection operation is first performed on the density distribution data. Specifically, the two-dimensional density matrix is ​​integrated along the row direction, and the density values ​​of each column of pixels are accumulated to generate a vertical density distribution curve.

[0145] Step b2: Perform bimodal detection on the longitudinal density distribution curve to identify the first density core region and the second density core region.

[0146] Bimodal detection refers to the algorithmic process of identifying two local maxima regions in a curve through derivative analysis. The first density core region can refer to the area of ​​dense crop growth corresponding to the first peak in the curve. The second density core region can refer to the area of ​​dense crop growth corresponding to the second peak in the curve.

[0147] In this embodiment, the longitudinal density distribution curve is first subjected to bimodal detection. The two maximum points are identified by calculating the zero point of the first derivative of the curve. Then, the maximum points are extended to both sides to the valley points to divide the first density core area and the second density core area.

[0148] Step b3: Perform a radial density scan operation within the first density core region and the second density core region to generate the first core point coordinate set and the second core point coordinate set.

[0149] The radial density scan operation refers to a calculation method that searches for local density extrema points along a radial path centered on a core point. The first core point coordinate set refers to the dataset of high-density point locations obtained by scanning within the first density core region. The second core point coordinate set refers to the dataset of high-density point locations obtained by scanning within the second density core region.

[0150] In this embodiment, a radial density scan operation is first performed within the first density core region, centered on the peak point, to radiate outwards at fixed angular intervals and search for local density maxima, generating a set of coordinates for the first core point. Simultaneously, the same operation is performed within the second density core region to generate a set of coordinates for the second core point.

[0151] Step b4: Merge the coordinate sets of the first core point and the second core point to form the coordinate set of the density core region.

[0152] In this embodiment, the first set of core point coordinates and the second set of core point coordinates are first input into the data fusion module, and then duplicate points are merged through coordinate deduplication processing, and finally the complete set of density core region coordinates is output.

[0153] Here's a specific example: When the harvester enters area C, the density distribution map is first longitudinally projected to generate a distribution curve with bimodal characteristics. Next, derivative analysis identifies the peak points at 2.1m and 4.3m longitudinally, dividing the area into a first core zone and a second core zone. Then, within the first core zone, a radial scan is performed with a center at 2.1m to obtain the coordinates of three core points, and within the second core zone, with a center at 4.3m, the coordinates of four core points are obtained. Finally, these seven coordinate points are merged to form the density core area coordinate set.

[0154] By executing steps b1 to b4, this embodiment of the application identifies bimodal features through longitudinal projection and accurately locates high-density points by combining radial scanning, providing precise crop distribution feature input for pose mapping.

[0155] In one possible embodiment, S14, generating a pose control signal based on the pose error compensation amount, includes:

[0156] Step 141: Input the pose error compensation amount into the servo signal converter, perform the control amount allocation operation, and generate the first height control amount and the first tilt angle control amount.

[0157] The servo signal converter can refer to a dedicated hardware module that decomposes spatial pose error into independent actuator control quantities. The control quantity allocation operation can refer to the calculation process of allocating vertical and rotational control quantities according to the orthogonal decomposition principle of degrees of freedom. The first height control quantity can refer to the original vertical displacement command value without considering the physical limitations of the actuator. The first tilt angle control quantity can refer to the original rotation angle command value without considering the motor torque limitations.

[0158] In this embodiment, the pose error compensation amount is first input to the servo signal converter, and the spatial compensation amount is decomposed into vertical direction component and rotation direction component through the decoupling allocation algorithm to generate the first height control amount and the first tilt angle control amount.

[0159] Step 142: Perform anti-saturation constraint operation on the first height control quantity and the first tilt angle control quantity to generate the second height control quantity and the second tilt angle control quantity.

[0160] Among them, anti-saturation constraint operation can refer to the process of preventing control commands from exceeding the physical limits of the actuator through a limiting algorithm. The second height control quantity can refer to the effective hydraulic cylinder extension / retraction control quantity after stroke constraint processing. The second tilt angle control quantity can refer to the effective motor rotation control quantity after torque constraint processing.

[0161] In this embodiment, the stroke limit parameters of the actuator are first read, and then the first height control quantity is limited to ensure that it does not exceed the maximum stroke of the hydraulic cylinder. At the same time, the first tilt angle control quantity is constrained by torque limit, and a second height control quantity and a second tilt angle control quantity that meet the physical constraints are generated.

[0162] Step 143: Based on the second height control quantity, generate the corresponding servo valve control signal; based on the second tilt angle control quantity, generate the pulse width modulation signal corresponding to the tilt angle using the preset electromechanical coupling equation.

[0163] Here, the servo valve control signal can refer to the analog current control command that drives the hydraulic proportional valve. The electromechanical coupling equation can refer to the differential equation model describing the relationship between motor torque and rotation angle.

[0164] In this embodiment, the second height control quantity is first input to the hydraulic control module, and a servo valve control signal for driving the hydraulic proportional valve is generated through a voltage-displacement conversion function. Simultaneously, the second tilt angle control quantity is input to the motor drive module, and the rotation control quantity is converted into a pulse width modulation signal with a corresponding duty cycle using an electromechanical coupling equation.

[0165] Step 144: Combine the servo valve control signal and the pulse width modulation signal to form the pose control signal.

[0166] Among them, the pulse width modulation signal can refer to the square wave duty cycle digital command that controls the speed of a DC motor.

[0167] In this embodiment, the servo valve control signal and the pulse width modulation signal are first processed for timing synchronization, and then encapsulated into data frames according to the control bus protocol, ultimately forming a pose control signal packet that simultaneously contains hydraulic execution instructions and motor drive instructions.

[0168] Here's a specific example: When the system generates a compensation amount requiring a 2cm height and a 5-degree tilt adjustment, the servo signal converter first decomposes it into a 2cm height control amount and a 5-degree tilt control amount. Next, it detects that the hydraulic cylinder's remaining stroke is only 3.5cm, constraining the height control amount to 2cm. Simultaneously, it detects that the motor can output torque to maintain the tilt control amount at 5 degrees. Then, the height control amount is converted into an 8mA hydraulic valve current signal, and the tilt control amount is converted into a 60% duty cycle PWM signal through electromechanical equations. Finally, the current signal and the PWM signal are encapsulated into a pose control signal and sent to the actuator.

[0169] By executing steps 141 to 144, this embodiment of the application generates safe and reliable execution instructions by intelligently allocating dual-channel control quantities and implementing physical constraints, ensuring that the posture adjustment system operates accurately and stably within the device's limit range.

[0170] Figure 2 This is a schematic diagram of the operation control system of a crop harvester provided in an embodiment of this application, as shown below. Figure 2 As shown, the system includes:

[0171] The acquisition module 21 is used to acquire density distribution data of the target object and terrain feature data of the area where the target object is located.

[0172] The mapping module 22 is used to map density distribution data and terrain feature data into target pose parameters of the controlled object based on a preset dynamic mapping model of density, terrain and pose. The target pose parameters include: height setting value and tilt angle setting value. The controlled object is a multi-degree-of-freedom actuator of a crop harvester.

[0173] Input module 23 is used to obtain the actual pose parameters of the controlled object, input the actual pose parameters and the target pose into the adaptive controller, and dynamically calculate the pose error compensation amount through the closed-loop control algorithm.

[0174] The drive module 24 is used to generate a pose control signal based on the pose error compensation amount, and drive the pose adjustment structure to perform pose calibration on the controlled object, so as to adjust the working height and tilt angle of the end effector installed at the front end of the multi-degree-of-freedom actuator.

[0175] Figure 2 The aforementioned operation control system for a crop harvester can execute... Figure 1 The implementation principle and technical effects of the crop harvester operation control method described in the illustrated embodiment will not be repeated here. The specific methods by which each module and unit of the crop harvester operation control system in the above embodiments perform operations have been described in detail in the embodiments related to this method, and will not be elaborated upon here.

[0176] In one possible design, Figure 2 The operation control system of a crop harvester in the illustrated embodiment can be implemented as a computing device, such as... Figure 3 As shown, the computing device may include a storage component 31 and a processing component 32.

[0177] The storage component 31 stores one or more computer instructions, wherein the one or more computer instructions are invoked and executed by the processing component 32.

[0178] The processing component 32 is used to perform the following processes: acquiring density distribution data of the target object and terrain feature data of the area where the target object is located. Based on a preset dynamic mapping model of density, terrain, and pose, the density distribution data and terrain feature data are mapped to target pose parameters of the controlled object. The target pose parameters include: height setpoint and tilt angle setpoint. The controlled object is a multi-degree-of-freedom actuator of a crop harvester. The actual pose parameters of the controlled object are acquired, and the actual pose parameters and target pose are input into an adaptive controller. The pose error compensation amount is dynamically calculated through a closed-loop control algorithm. Based on the pose error compensation amount, a pose control signal is generated to drive the pose adjustment structure to perform pose calibration on the controlled object, so as to adjust the working height and tilt angle of the end effector installed at the front end of the multi-degree-of-freedom actuator.

[0179] The processing component 32 may include one or more processors to execute computer instructions to complete all or part of the steps in the above-described method. Alternatively, the processing component may be implemented as one or more application-specific integrated circuits (ASICs), digital signal processors (DSPs), digital signal processing devices (DSPDs), programmable logic devices (PLDs), field-programmable gate arrays (FPGAs), controllers, microcontrollers, microprocessors, or other electronic components to perform the above-described method.

[0180] Storage component 31 is configured to store various types of data to support operations at the terminal. The storage component can be implemented from any type of volatile or non-volatile storage device or a combination thereof, such as Random Access Memory (RAM), Static Random-Access Memory (SRAM), Electrically Erasable Programmable Read Only Memory (EEPROM), Erasable Programmable Read Only Memory (EPROM), Programmable Read Only Memory (PROM), Read Only Memory (ROM), magnetic storage, flash memory, magnetic disk, or optical disk.

[0181] Of course, computing devices may also include other components, such as input / output interfaces, display components, communication components, etc.

[0182] Input / output interfaces provide interfaces between processing components and peripheral interface modules, which can be output devices, input devices, etc.

[0183] The communication components are configured to facilitate wired or wireless communication between computing devices and other devices.

[0184] The computing device can be a physical device or an elastic computing host provided by a cloud computing platform. In this case, the computing device can refer to a cloud server, and the aforementioned processing components, storage components, etc., can be basic server resources rented or purchased from the cloud computing platform.

[0185] This application also provides a computer storage medium storing a computer program, which, when executed by a computer, can perform the above-described functions. Figure 1 The embodiment shown illustrates an operation control method for a crop harvester.

[0186] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0187] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Those skilled in the art can understand and implement this without any creative effort.

[0188] Through the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus necessary general-purpose hardware platforms, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solutions, in essence or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods described in the various embodiments or some parts of the embodiments.

[0189] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application.

Claims

1. A method for controlling the operation of a crop harvester, characterized in that, include: Acquire density distribution data of the target object and terrain feature data of the area where the target object is located. The terrain feature data includes surface elevation, slope angle and curvature generated by three-dimensional scanning. Based on a preset dynamic mapping model of density, terrain and pose, the density distribution data and the terrain feature data are mapped to the target pose parameters of the controlled object. The target pose parameters include: height setting value and tilt angle setting value. The controlled object is a multi-degree-of-freedom actuator of a crop harvester. The actual pose parameters of the controlled object are obtained, and the actual pose parameters and the target pose are input into the adaptive controller. The pose error compensation amount is dynamically calculated through a closed-loop control algorithm. The adaptive controller includes a fuzzy proportional-integral-derivative controller. Based on the posture error compensation amount, a posture control signal is generated to drive the posture adjustment structure to perform posture calibration on the controlled object, so as to adjust the working height and tilt angle of the end effector installed at the front end of the multi-degree-of-freedom actuator. The posture adjustment structure is a posture execution device that includes a hydraulic valve group and an electric push rod. The step of inputting the actual pose parameters and the target pose into the adaptive controller, and dynamically calculating the pose error compensation amount through a closed-loop control algorithm, includes: In the adaptive controller, the actual height value and the actual tilt angle value are obtained from the actual pose parameters; The set height value and the actual height value are tracked at a set point to generate a height deviation signal; the set tilt angle value and the actual tilt angle value are tracked at the set point to generate a tilt angle deviation signal. The multi-parameter coupling compensation module of the adaptive controller, combined with the dynamic transfer function and acceleration residual calculation technology of the multi-degree-of-freedom actuator, performs disturbance estimation on the height deviation signal and the tilt angle deviation signal to obtain the pose disturbance component containing the height disturbance component and the tilt angle disturbance component; wherein, the dynamic transfer function explicitly characterizes the inertial and elastic coupling mechanism of the height and tilt angle channels through the asymmetric coupling matrix. The formula for the dynamic transfer function is: ； in, Input force for height channel control, The tilt channel control input torque is... For the equivalent mass in the height direction, The inertial coupling coefficient is the inclination angle and height. The inertial coupling coefficient for height and tilt angle. The moment of inertia is the rotational inertia in the tilt direction. For the linear acceleration of the end effector, For the end effector angular acceleration, The damping coefficient is in the height direction. The damping coupling coefficient for height and tilt angle. The damping coupling coefficient is the inclination angle and height. The damping coefficient is in the tilt direction. For the rate of change of height, For the angular velocity of the tilt angle change, Here is the stiffness coefficient in the height direction. This is the stiffness coefficient in the tilt direction. This represents the actual height of the end effector. This represents the actual tilt angle of the end effector. External disturbances to the high-altitude channel External disturbances to the tilt channel; The height disturbance component and the tilt disturbance component are compared with the corresponding preset stability boundary conditions to obtain the dynamically adjusted height channel gain coefficient and tilt channel gain coefficient. Using a closed-loop control algorithm, the height deviation signal and the tilt deviation signal are processed based on the height channel gain coefficient and the tilt channel gain coefficient to generate the pose error compensation amount; The dynamic mapping model based on preset density, terrain, and pose maps the density distribution data and the terrain feature data to the target pose parameters of the controlled object, including: Spatial density focusing operation is performed on the density distribution data to generate a set of coordinates for the density core region; curvature decomposition operation is performed on the terrain feature data to generate a terrain gradient tensor. Based on the density core region coordinate set and the terrain gradient tensor, the initial height value is generated using the height mapping function in the dynamic mapping model, which is used to characterize the relationship between density, terrain and height. Based on the density core region coordinate set and the terrain gradient tensor, the initial value of the dip angle is generated using the dip angle mapping function in the dynamic mapping model, which is used to characterize the relationship between density, terrain and dip angle. Based on the compensation coefficient corresponding to the growth stage of the target object, a compensation operation is performed on the initial height value to generate a height setting value. Based on the influence of the ambient wind speed on the target object, an adaptive compensation operation is performed on the initial tilt angle value to generate a tilt angle setting value. The target pose parameters are formed by combining the height setting value and the tilt angle setting value. Based on the height channel gain coefficient and the tilt channel gain coefficient, the height deviation signal and the tilt deviation signal are processed to generate a pose error compensation amount, including: Using a closed-loop control algorithm, the height deviation signal is time-varying weighted by the height channel gain coefficient, and the tilt deviation signal is time-varying weighted by the tilt channel gain coefficient, to generate a weighted height deviation signal and a weighted tilt deviation signal. Dead zone crossing processing is performed on the weighted height deviation signal and the weighted tilt angle deviation signal to generate height compensation component and tilt angle compensation component; The height compensation component and the tilt compensation component are combined to form the pose error compensation amount.

2. The method according to claim 1, characterized in that, The step of performing dead-zone crossing processing on the weighted height deviation signal and the weighted tilt angle deviation signal to generate height compensation components and tilt angle compensation components includes: Mechanical vibration data is collected by an inertial measurement unit installed on the multi-degree-of-freedom actuator, and the peak vibration energy is calculated based on the mechanical vibration data. When the peak vibration energy exceeds the preset vibration threshold, the weighted height deviation signal and the weighted tilt angle deviation signal are attenuated using a preset vibration suppression coefficient to obtain the height correction value and the tilt angle correction value. Based on the motion state of the multi-degree-of-freedom actuator, dynamically calculate the height dead zone threshold and tilt angle dead zone threshold; In the nonlinear dead zone compensation process, if the height correction value is greater than the height dead zone threshold, a height compensation component is generated based on the height correction value using a preset height compensation formula. If the tilt angle correction value is greater than the tilt angle dead zone threshold, a tilt angle compensation component is generated based on the tilt angle correction value using a preset tilt angle compensation formula.

3. The method according to claim 1, characterized in that, The step of performing a spatial density focusing operation on the density distribution data to generate a set of coordinates for the density core region includes: Perform a longitudinal density projection operation on the density distribution data to generate a longitudinal density distribution curve; Perform bimodal detection on the longitudinal density distribution curve to identify the first density core region and the second density core region; A radial density scan operation is performed within the first density core region and the second density core region to generate a first core point coordinate set and a second core point coordinate set. The first core point coordinate set and the second core point coordinate set are merged to form the density core region coordinate set.

4. The method according to claim 1, characterized in that, The generation of the pose control signal based on the pose error compensation amount includes: The pose error compensation amount is input into the servo signal converter to perform a control amount allocation operation and generate a first height control amount and a first tilt angle control amount. Perform anti-saturation constraint operation on the first height control value and the first tilt angle control value to generate the second height control value and the second tilt angle control value. Based on the second height control quantity, a corresponding servo valve control signal is generated; based on the second tilt angle control quantity, a pulse width modulation signal corresponding to the tilt angle is generated using a preset electromechanical coupling equation. The servo valve control signal and the pulse width modulation signal are combined to form a pose control signal.

5. An operation control system for a crop harvester, characterized in that, A method for controlling the operation of a crop harvester as described in claim 1 includes: The acquisition module is used to acquire density distribution data of the target object and terrain feature data of the area where the target object is located; The mapping module is used to map the density distribution data and the terrain feature data into target pose parameters of the controlled object based on a preset dynamic mapping model of density, terrain and pose. The target pose parameters include: height setting value and tilt angle setting value. The controlled object is a multi-degree-of-freedom actuator of a crop harvester. The input module is used to obtain the actual pose parameters of the controlled object, input the actual pose parameters and the target pose into the adaptive controller, and dynamically calculate the pose error compensation amount through the closed-loop control algorithm. The drive module is used to generate a pose control signal based on the pose error compensation amount, and drive the pose adjustment structure to perform pose calibration on the controlled object, so as to adjust the working height and tilt angle of the end effector installed at the front end of the multi-degree-of-freedom actuator.

6. A computing device, characterized in that, It includes a processing component and a storage component; the storage component stores one or more computer instructions; the one or more computer instructions are invoked and executed by the processing component to implement the operation control method of a crop harvester as described in any one of claims 1 to 4.

7. A computer storage medium, characterized in that, The device contains a computer program that, when executed by a computer, implements an operation control method for a crop harvester as described in any one of claims 1 to 4.

Images