A variable load adaptive trajectory tracking control method for a dual-steering wheel omnidirectional AGV

Abstract

This invention provides a variable load adaptive trajectory tracking control method for a dual-steering-wheel omnidirectional AGV, relating to the field of AGV motion control technology. The method includes: acquiring vehicle state data and load state data of the AGV to construct state feature data; identifying load changes based on the state feature data, determining motion response change characteristics, and updating the vehicle motion state model accordingly to obtain real-time motion state information; determining trajectory tracking deviation data based on the real-time motion state information and target trajectory information, and performing state correlation analysis in conjunction with motion response change characteristics to obtain control adjustment requirement information; adaptively adjusting trajectory tracking control parameters accordingly to generate steering control and drive control quantities for each steering wheel; acquiring vehicle state data after executing motion control, comparing it with the target trajectory to update the deviation data, and achieving closed-loop adaptive control. This invention can adjust control parameters online according to real-time load changes, effectively improving trajectory tracking accuracy under variable load conditions.

Description

Technical Field

[0001] This invention relates to the field of AGV motion control technology, specifically to a variable load adaptive trajectory tracking control method for a dual-steering wheel omnidirectional AGV. Background Technology

[0002] Dual-steering-wheel omnidirectional AGVs (Automated Guided Vehicles), as intelligent material handling equipment with omnidirectional mobility, can achieve multi-degree-of-freedom movements such as forward, backward, lateral, diagonal, and rotation through the coordinated drive of two independent steering wheels. They possess irreplaceable application advantages in scenarios such as heavy-duty handling, precision assembly, and logistics in confined spaces. However, the core technical challenge faced by this type of AGV in actual operation lies in the dynamic changes in the vehicle's mass and center of gravity caused by loading and unloading goods. These changes significantly alter the vehicle's inertial parameters and tire-ground adhesion characteristics, resulting in nonlinear interference with the vehicle's acceleration and steering responses. Consequently, the control accuracy of trajectory tracking controllers designed based on fixed-parameter models severely decreases under variable load conditions, making it difficult to simultaneously maintain tracking accuracy and driving stability under different load states.

[0003] In the prior art, a trajectory tracking control method for an unmanned transport vehicle with a variable center of mass, disclosed in CN113467478A, establishes a dynamic and kinematic model of the unmanned transport vehicle, designs a global asymptotic tracking controller based on the Lyapunov direct method, and uses an adaptive particle swarm optimization algorithm to offline optimize and adjust the fixed control parameters of the controller to achieve trajectory tracking control under variable center of mass conditions. However, the control parameter optimization in this scheme is an offline one-time tuning, and the optimized control parameters remain fixed during operation. When the load of the transport vehicle changes in real time during loading and unloading, the pre-tuned control parameters cannot be adaptively adjusted online according to the real-time dynamic changes in the load. In addition, this scheme does not involve real-time identification and monitoring of load changes, nor does it establish a dynamic response correlation mechanism between load changes and controller parameter adjustments. As a result, when the load changes frequently during operation, the controller struggles to maintain high-precision trajectory tracking performance, and the problems of trajectory deviation and tracking error accumulation remain prominent.

[0004] The information disclosed in the background section is only intended to enhance the understanding of the background of this disclosure, and therefore may include information that does not constitute prior art known to those skilled in the art. Summary of the Invention

[0005] The purpose of this invention is to provide a variable load adaptive trajectory tracking control method for a dual-steering wheel omnidirectional AGV to solve the problems mentioned in the background art.

[0006] To achieve the above objectives, the present invention provides a variable load adaptive trajectory tracking control method for a dual-steering wheel omnidirectional AGV, the specific steps of which include: Step 1: Acquire vehicle status data and load status data during AGV operation. The vehicle status data includes vehicle pose data, dual steering wheel angle data, and drive wheel motion data. The load status data includes load mass and load change. Based on the vehicle status data and load status data, construct status feature data corresponding to the current operating condition. Step 2: Based on the state feature data, identify the current load change of the AGV, determine the motion response change characteristics caused by the load change, and update the pre-built vehicle motion state model based on the motion response change characteristics to obtain real-time motion state information that matches the current load state. Step 3: Based on the real-time motion state information and the preset target trajectory information, determine the trajectory tracking deviation data of the AGV, and combine the motion response change characteristics to perform state correlation analysis on the trajectory tracking deviation data to obtain control adjustment requirements information characterizing the trajectory tracking performance under the current load condition; Step 4: Based on the control adjustment requirement information, adaptively adjust the trajectory tracking control parameters of the AGV to generate steering wheel control parameters that match the current load state, and then determine the steering control quantity and drive control quantity corresponding to the two steering wheels. Step 5: Perform motion control on the AGV based on the steering control and drive control quantities, and acquire the vehicle status data after control in real time. Compare the vehicle status data with the target trajectory information and update the trajectory tracking deviation data to achieve adaptive trajectory tracking control of the AGV under different load conditions.

[0007] Furthermore, step 1 specifically includes: By using a positioning detection module, a steering wheel status detection module, and a load detection module installed on the AGV body, vehicle pose data, dual steering wheel motion data, and load status data during AGV operation are obtained, respectively. Among them, the positioning detection module obtains the AGV's current position coordinates and vehicle heading angle in the global coordinate system to form vehicle pose data. The vehicle pose data is composed of the AGV's lateral position coordinates, longitudinal position coordinates, and vehicle heading angle at the current moment. The steering wheel status detection module collects the steering angle and driving speed of the two steering wheels respectively to form dual steering wheel motion data. The dual steering wheel motion data is composed of the steering angle of the first steering wheel at the current moment, the steering angle of the second steering wheel at the current moment, the driving speed of the first steering wheel at the current moment, and the driving speed of the second steering wheel at the current moment.

[0008] Furthermore, the load detection module obtains the current load status information of the AGV and generates load status data based on the characteristic parameters that affect the vehicle's motion performance. The load status data is composed of the AGV's load mass at the current moment and the current load change. The current load change is used to characterize the load change between adjacent sampling moments, and it is determined by the difference between the load mass obtained at the current moment and the load mass obtained at the previous moment. The vehicle pose data, dual steering wheel motion data, and load status data are processed for time synchronization, and the status feature data corresponding to the current operating condition is constructed according to a preset arrangement rule.

[0009] Furthermore, step 2 specifically includes: S2.1: Based on the load status data in the state feature data, identify the current load change of the AGV. The load change includes the direction and rate of change of the load mass. Based on the direction and rate of change of the load mass, determine the vehicle motion response change characteristics caused by the load change. The motion response change characteristics include acceleration response change characteristics and steering response change characteristics. S2.2: Based on the acceleration response change characteristics, update the velocity state equation in the pre-constructed vehicle motion state model to obtain a velocity state update equation that matches the current load state; S2.3: Based on the steering response change characteristics, update the attitude state equation in the pre-constructed vehicle motion state model to obtain an attitude state update equation that matches the current load state. S2.4: Based on the velocity state update equation and attitude state update equation, the vehicle pose data and dual steering wheel motion data in the state feature data at the current moment are used as inputs to the updated vehicle motion state model to calculate real-time motion state information that matches the current load state. The real-time motion status information includes at least the predicted lateral speed, predicted longitudinal speed, and predicted yaw rate of the AGV under the current load condition.

[0010] Furthermore, step 3 specifically includes: S3.1: Based on the real-time motion status information and the preset target trajectory information, determine the trajectory tracking deviation data of the AGV in the global coordinate system. The trajectory tracking deviation data includes lateral position deviation, longitudinal position deviation and heading angle deviation. S3.2: Obtain the acceleration response change characteristics and steering response change characteristics from the motion response change characteristics, perform state correlation analysis on the acceleration response change characteristics with the lateral position deviation and longitudinal position deviation, and determine the first influencing factor of the acceleration response deviation caused by load change on the position tracking performance; S3.3: Perform state correlation analysis between the steering response change characteristics and the heading angle deviation to determine the second influencing factor of the steering response deviation caused by load change on the heading tracking performance; S3.4: Based on the first influence factor and the second influence factor, the trajectory tracking deviation data is corrected by load influence compensation to obtain effective tracking deviation data after motion response feature correction; S3.5: Based on the effective tracking deviation data and combined with the real-time motion state information at the current moment, generate control adjustment requirement information characterizing the trajectory tracking performance under the current load condition. The control adjustment requirement information includes at least the position loop control adjustment requirement and the heading loop control adjustment requirement.

[0011] Furthermore, the logic for determining the steering wheel control parameters in step 4 is as follows: Based on the position loop control adjustment demand and the heading loop control adjustment demand in the control adjustment demand information, determine the position loop control parameter adjustment and the heading loop control parameter adjustment respectively. Based on the adjustment amount of the position loop control parameters, the proportional gain coefficient and integral gain coefficient of the position loop controller are adaptively adjusted to generate position loop control gain parameters that match the current load state. Based on the adjustment amount of the heading loop control parameters, the proportional gain coefficient and derivative gain coefficient of the heading loop controller are adaptively adjusted to generate heading loop control gain parameters that match the current load state. The position loop control gain parameter and the heading loop control gain parameter are input to the trajectory tracking controller. The trajectory tracking controller calculates the control quantity based on the real-time motion state information and effective tracking deviation data at the current moment, and generates a vehicle control quantity that matches the current load state. The vehicle control quantity includes the vehicle target linear velocity and the vehicle target angular velocity. The steering wheel control parameters are characterized by vehicle control variables.

[0012] Furthermore, the logic for determining the steering control quantity and the drive control quantity is as follows: The target linear velocity and target angular velocity of the vehicle are input into the dual-steering wheel inverse kinematics module. Based on the geometric arrangement parameters of the dual steering wheels, the inverse kinematics calculation is performed, and the target linear velocity and target angular velocity of the vehicle are distributed to the two steering wheels, generating the steering control and drive control quantities corresponding to the first steering wheel and the steering control and drive control quantities corresponding to the second steering wheel, respectively.

[0013] Furthermore, real-time acquisition of vehicle status data after control is performed, specifically: The steering control and drive control quantities corresponding to the first steering wheel are sent to the steering drive motor and the travel drive motor of the first steering wheel, and the steering control and drive control quantities corresponding to the second steering wheel are sent to the steering drive motor and the travel drive motor of the second steering wheel, driving the AGV to execute motion according to the motion command calculated in the current control cycle. At the next sampling moment after the motion is performed, the updated vehicle status data is obtained in real time through the positioning detection module, the steering wheel status detection module and the load detection module. The updated vehicle status data includes updated vehicle pose data, updated dual steering wheel motion data and updated load status data.

[0014] Furthermore, the updated vehicle pose data is compared with the preset target trajectory information, and the trajectory tracking deviation data at the current moment is recalculated to obtain the updated trajectory tracking deviation data. Determine whether the updated trajectory tracking deviation data meets the preset trajectory tracking accuracy threshold condition; If the trajectory tracking accuracy threshold condition is met, the current control parameters remain unchanged, and the updated vehicle status data is used as the initial status data for the next control cycle, returning to step 1; If the trajectory tracking accuracy threshold condition is not met, the updated trajectory tracking deviation data is used as the initial deviation data for the next control cycle, and the updated vehicle state data is used as the initial state data for the next control cycle. Then, return to step 2 and re-execute load change recognition, motion response feature extraction, adaptive adjustment of control parameters, and steering wheel control quantity calculation until the trajectory tracking deviation of the AGV under different load conditions meets the trajectory tracking accuracy threshold condition, thereby realizing the closed-loop adaptive trajectory tracking control of the AGV.

[0015] Compared with the prior art, the beneficial effects of the present invention are: This invention acquires vehicle status data and load status data during AGV operation, identifies changes in load mass and their rate of change, and combines the acceleration response and steering response characteristics caused by load changes to dynamically update the vehicle motion state model. This results in obtaining real-time motion state information that matches the current load state, enabling the vehicle motion state to be adjusted in real time with load changes and improving the accuracy of vehicle motion state prediction under variable load conditions. The present invention also performs state correlation analysis on the motion response change characteristics caused by load change and the trajectory tracking deviation to determine the influence factors of acceleration response change and steering response change on position tracking performance and heading tracking performance, respectively, and performs load influence compensation correction on trajectory tracking deviation accordingly to obtain effective tracking deviation data, thereby reducing the interference of motion response disturbance caused by load change on trajectory tracking control and avoiding over-adjustment of the controller for non-control performance factors. This invention also generates control adjustment requirements for position loop and heading loop based on effective tracking deviation data, and adaptively adjusts the control gain parameters of position loop and heading loop respectively, so that the controller can dynamically match the corresponding control parameters according to the current load conditions. Then, through the inverse kinematics solution of the dual steering wheels, the vehicle control quantity is converted into the steering control quantity and drive control quantity corresponding to the two steering wheels, so as to realize the coordinated control of the dual steering wheels and improve the trajectory tracking accuracy and motion stability of AGV under different load conditions. This invention also acquires updated vehicle status data in real time after motion control is executed, recalculates trajectory tracking deviation, and determines whether to maintain the current control parameters or re-execute load change recognition, motion status update, and adaptive adjustment of control parameters based on the trajectory tracking accuracy judgment result. This enables the control process to form a closed-loop feedback, realize continuous response to load changes and trajectory tracking status, and enhance the trajectory tracking adaptability of AGV in frequent loading and unloading and variable load operation. Attached Figure Description

[0016] Figure 1 This is a schematic diagram of the overall method flow of the present invention; Figure 2 This is a comparison chart of the maximum lateral position deviation between the embodiments of the present invention and the traditional control method under different loads; Figure 3 This is a comparison chart of the maximum heading angle deviation between the embodiments of the present invention and the traditional control method under different loads. Detailed Implementation

[0017] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to specific embodiments.

[0018] It should be noted that, unless otherwise defined, the technical or scientific terms used in this invention should have the ordinary meaning understood by one of ordinary skill in the art to which this invention pertains. The terms "first," "second," and similar terms used in this invention do not indicate any order, quantity, or importance, but are merely used to distinguish different components. Terms such as "comprising" or "including" mean that the element or object preceding the word encompasses the elements or objects listed following the word and their equivalents, without excluding other elements or objects. Terms such as "connected" or "linked" are not limited to physical or mechanical connections, but can include electrical connections, whether direct or indirect. Terms such as "upper," "lower," "left," and "right" are used only to indicate relative positional relationships; when the absolute position of the described object changes, the relative positional relationship may also change accordingly.

[0019] Example: Please refer to Figures 1-3 This invention provides a variable load adaptive trajectory tracking control method for a dual-steering wheel omnidirectional AGV, the specific steps of which include: Step 1: Acquire vehicle status data and load status data during AGV operation. The vehicle status data includes vehicle pose data, dual steering wheel angle data, and drive wheel motion data. The load status data includes load mass and load change. Based on the vehicle status data and load status data, construct status feature data corresponding to the current operating condition. In this embodiment, step 1 specifically includes: By using a positioning detection module, a steering wheel status detection module, and a load detection module installed on the AGV body, vehicle pose data, dual steering wheel motion data, and load status data during AGV operation are acquired, respectively. Specifically, the positioning detection module acquires the AGV's current position coordinates and heading angle in the global coordinate system, forming vehicle pose data. This vehicle pose data is composed of the AGV's lateral position coordinates, longitudinal position coordinates, and heading angle at the current moment. ; In the formula, Indicates the current state of the AGV. Corresponding vehicle pose data; Indicates the current state of the AGV. The corresponding horizontal position coordinates; Indicates the current state of the AGV. The corresponding vertical position coordinates; Indicates the current state of the AGV. The corresponding vehicle heading angle; The steering wheel state detection module collects the steering angle and driving speed of the two steering wheels respectively, forming dual steering wheel motion data. The dual steering wheel motion data is composed of the steering angle of the first steering wheel at the current moment, the steering angle of the second steering wheel at the current moment, the driving speed of the first steering wheel at the current moment, and the driving speed of the second steering wheel at the current moment. ; In the formula, Indicates the current state of the AGV. Corresponding dual steering wheel motion data; , These represent the first and second steering wheels at the current moment, respectively. The corresponding steering angle; , These represent the first and second steering wheels at the current moment, respectively. The corresponding drive speed.

[0020] Furthermore, the load detection module acquires the current load status information of the AGV and generates load status data based on the characteristic parameters that affect the vehicle's motion performance. This load status data is composed of the AGV's load mass at the current moment and the current load change. The current load change characterizes the load change between adjacent sampling moments and is determined by the difference between the load mass acquired at the current moment and the load mass acquired at the previous moment. ; In the formula, Indicates the current state of the AGV. Corresponding load status data; Indicates the current state of the AGV. The load-bearing capacity; This represents the current load change, and its calculation method is as follows: ; in, This indicates the load-bearing mass obtained at the previous moment; The vehicle pose data, dual steering wheel motion data, and load status data are processed for time synchronization, and the status feature data corresponding to the current operating condition is constructed according to a preset arrangement rule.

[0021] The AGV body is equipped with a positioning detection module, a steering wheel status detection module, and a load detection module. The positioning detection module preferably uses a laser SLAM navigation system or a differential GPS positioning system, fixedly installed on the top or center of the AGV body, to acquire the AGV's position coordinates and heading angle in the global coordinate system in real time. The steering wheel status detection module includes an angle sensor and a speed sensor respectively installed on the first and second steering wheels. The angle sensor is preferably an absolute rotary encoder, installed on the steering shaft of the steering wheel, to acquire the steering angle of the steering wheel in real time. The speed sensor is preferably an incremental encoder, installed on the output shaft of the drive motor or at the hub of the steering wheel, to acquire the drive speed of the steering wheel in real time. The load detection module uses a load sensing array composed of multiple pressure sensors. This load sensing array is located at the connection between the AGV body and the cargo platform. The pressure sensors are evenly distributed on the load-bearing surface of the body according to a preset array layout, to acquire the load distribution force value at each sensing point in real time, thereby obtaining the current load status information of the AGV.

[0022] During the actual operation of the AGV, the positioning and detection module collects data on the AGV's position at the current moment according to a preset sampling period. Horizontal position coordinates Vertical position coordinates and vehicle heading angle These three parameters together constitute the vehicle pose data at the current moment. The steering wheel status detection module synchronously collects the status data of the first steering wheel at the current moment. steering angle The second steering wheel at the current moment steering angle Simultaneously, the first steering wheel's position at the current moment is collected. drive speed The second steering wheel at the current moment drive speed These four parameters together constitute the motion data of the dual steering wheels at the current moment. The load detection module synchronously collects the load distribution force values ​​at each sensor point, and calculates the AGV's position at the current moment based on the installation coordinates of each sensor point and the load force value it bears. load-bearing capacity At the same time, the load-bearing mass acquired at the current moment The load-bearing mass obtained from the previous moment Perform the difference calculation to obtain the current load change. Bearing capacity and current load change Together they constitute the load state data at the current moment. .

[0023] After collecting the above three sets of data, the vehicle pose data is synchronized using the sampling time of the positioning detection module as the synchronization reference. Dual steering wheel motion data and load status data Time synchronization is performed. Specifically, if there is a time deviation between the sampling time of the steering wheel state detection module and the load detection module and the sampling time of the positioning detection module, interpolation is used to align all data sets to the same time reference point. After time synchronization is completed, the synchronized vehicle pose data, dual steering wheel motion data, and load state data are fused according to a preset arrangement rule to form a multi-dimensional state feature vector. This multi-dimensional state feature vector is the state feature data corresponding to the current operating condition and serves as the input data for load change identification and motion state update in subsequent steps. After each sampling cycle, the load mass at the current moment is... Stored in a cache unit for use in the next sampling period when calculating load changes.

[0024] Step 2: Based on the state feature data, identify the current load change of the AGV, determine the motion response change characteristics caused by the load change, and update the pre-built vehicle motion state model based on the motion response change characteristics to obtain real-time motion state information that matches the current load state. In this embodiment, step 2 specifically includes: S2.1: Based on the load status data in the state feature data, identify the current load change of the AGV. The load change includes the direction and rate of change of the load mass. Based on the direction and rate of change of the load mass, determine the vehicle motion response change characteristics caused by the load change. The motion response change characteristics include acceleration response change characteristics and steering response change characteristics. In step S2.1, the current load change of the AGV is identified based on the load status data in the status feature data. Specifically, the load capacity is extracted from the status feature data. and current load change The current load change Compare with zero: when When the value is greater than zero, the AGV is determined to be in a loading state, and the load-bearing capacity increases; when... When the value is less than zero, the AGV is determined to be in an unloading state, and the load capacity is reduced; when... When the value equals zero, the AGV is determined to be in a constant load state, with the load-bearing mass remaining unchanged. Simultaneously, the current load change is recorded. The absolute value divided by the sampling period The rate of load change is obtained. Based on this, and according to the direction and rate of change of the load mass, combined with the vehicle's current speed, the characteristics of the vehicle's motion response caused by the load change are determined. Specifically, the acceleration response characteristics are determined by the longitudinal acceleration change caused by the load change. The characterization, its calculation formula is: ; in, ; In the formula, The change in longitudinal acceleration caused by load variation, i.e., the characteristic of acceleration response variation; This represents the reference longitudinal acceleration generated by the driving force at the current moment; This represents the resultant longitudinal force acting on the AGV. The physical meaning of this formula is: Under the premise that the longitudinal driving force remains constant, a sudden change in the load-bearing mass will cause a corresponding jump in the longitudinal acceleration. That is, used to quantify the magnitude of the jump, its sign and... The signs are opposite, meaning that the longitudinal acceleration decreases when loaded and increases when unloaded.

[0025] Steering response variation characteristics are the changes in yaw acceleration caused by load variations. The characterization, its calculation formula is: ; In the formula, This represents the change in yaw acceleration caused by load variation, i.e., the characteristic of steering response variation; The lateral acceleration of the AGV is acquired by the inertial measurement unit. and These are the lateral and longitudinal coordinates of the load centroid in the vehicle coordinate system at the current moment; This represents the lateral coordinate of the load's center of mass in the vehicle coordinate system at the previous moment; This represents the moment of inertia of the AGV relative to the vertical axis of its center of gravity in an unloaded state, and is a factory-calibrated parameter. The physical meaning of this formula is that when the load's center of gravity shifts to one side of the vehicle's longitudinal axis, an additional yaw moment is generated, which causes a change in the yaw acceleration. At the same time, the abrupt change in the position of the load's center of gravity (i.e., the sudden movement of the load's position on the vehicle body) will also have an instantaneous impact on the yaw acceleration.

[0026] S2.2: Based on the acceleration response change characteristics, update the velocity state equation in the pre-constructed vehicle motion state model to obtain a velocity state update equation that matches the current load state; The initial form of the velocity state equation in the pre-constructed vehicle motion state model is as follows: ; in, This represents the longitudinal acceleration of the AGV's center of mass; This indicates the total mass of the AGV in its unloaded state, which is a parameter calibrated at the vehicle's factory. This indicates the longitudinal resultant force acting on the AGV, which is calculated from the driving forces of the two steering wheels; The lateral velocity of the AGV is obtained by an inertial measurement unit located on the vehicle body. The yaw rate of the AGV is also obtained by the inertial measurement unit. When the load changes, the overall equivalent mass of the AGV becomes... The mass-related terms in the velocity state equation need to be updated accordingly, and the longitudinal acceleration change calculated in step S2.1 also needs to be introduced. As a load mutation compensation term, the updated velocity state equation is: ; In the formula, That is, the sum of the unloaded mass and the current load mass.

[0027] S2.3: Based on the steering response change characteristics, update the attitude state equation in the pre-constructed vehicle motion state model to obtain an attitude state update equation that matches the current load state. The initial form of the attitude state equation in the pre-constructed vehicle motion state model is as follows: ; in, This represents the yaw acceleration of the AGV; This represents the moment of inertia of the AGV relative to the vertical axis of its center of mass when it is unloaded. The yaw moment of the AGV is obtained by multiplying the actual lateral force of each of the two steering wheels by their respective longitudinal and lateral distances to the center of mass of the AGV, and then calculating by vector superposition. The actual lateral force of each steering wheel is collected in real time by wheel force sensors installed at the hub or steering knuckle of each steering wheel.

[0028] When the load changes, the overall moment of inertia of the AGV becomes Meanwhile, the displacement of the load's center of mass will affect the yaw moment, and the inertia-related terms in the attitude state equation need to be updated accordingly. In addition, the change in yaw acceleration calculated in step S2.1 also needs to be introduced. As a load offset compensation term, the updated attitude state equation is: ; in, The current overall rotational inertia of the AGV is calculated as follows: ; and These are the lateral and longitudinal coordinates of the load centroid in the vehicle coordinate system, respectively, which are calculated based on the weighted average of the load distribution force values ​​at each sensing point of the load sensing array and the installation position coordinates of each sensing point. The yaw moment acting on the AGV is calculated by converting the lateral forces of the two steering wheels through lever arms. This is the moment of inertia of the AGV relative to the vertical axis of its center of mass when it is unloaded; it is a factory-calibrated parameter.

[0029] S2.4: Based on the velocity state update equation and attitude state update equation, the vehicle pose data and dual steering wheel motion data in the state feature data at the current moment are used as inputs to the updated vehicle motion state model to calculate real-time motion state information that matches the current load state. The real-time motion status information includes at least the predicted lateral speed, predicted longitudinal speed, and predicted yaw rate of the AGV under the current load condition.

[0030] In step S2.4, based on the velocity state update equation and attitude state update equation, the vehicle pose data and dual steering wheel motion data in the current state feature data are used as inputs to the updated vehicle motion state model to calculate real-time motion state information matching the current load state. Specifically, the vehicle heading angle in the current vehicle pose data is used... and the drive speed of each steering wheel in the motion data of the dual steering wheels , and the steering angle of each steering wheel , Substituting the updated velocity and attitude state equations, the predicted lateral velocity of the AGV under the current load state is obtained by performing state prediction calculations using numerical integration. Predicting longitudinal velocity And predicting yaw rate The predicted lateral velocity and predicting longitudinal velocity The predicted yaw rate is the velocity component of the AGV in the vehicle coordinate system. Let be the angular velocity of the AGV around the vertical axis. The real-time motion state information calculated above will be used as input data for calculating the trajectory tracking deviation in step 3.

[0031] Step 3: Based on the real-time motion state information and the preset target trajectory information, determine the trajectory tracking deviation data of the AGV, and combine the motion response change characteristics to perform state correlation analysis on the trajectory tracking deviation data to obtain control adjustment requirements information characterizing the trajectory tracking performance under the current load condition; In this embodiment, step 3 specifically includes: S3.1: Based on the real-time motion status information and the preset target trajectory information, determine the trajectory tracking deviation data of the AGV in the global coordinate system. The trajectory tracking deviation data includes lateral position deviation, longitudinal position deviation and heading angle deviation. In step S3.1, the trajectory tracking deviation data of the AGV in the global coordinate system is determined based on the real-time motion state information and the preset target trajectory information. Specifically, the predicted lateral velocity, predicted longitudinal velocity, and predicted yaw rate of the AGV at the current moment are obtained from the real-time motion state information calculated in step 2. Combined with the vehicle pose calculated in the previous control cycle, the estimated actual pose of the AGV in the global coordinate system at the current moment is calculated through integration. The estimated actual pose includes the actual lateral position. Actual longitudinal position and actual heading angle Simultaneously, the desired pose corresponding to the current moment is extracted from the preset target trajectory information, the desired pose including the desired lateral position. Desired vertical position and expected heading angle The trajectory tracking deviation data in the global coordinate system is obtained by subtracting the actual pose estimate from the expected pose value item by item. The specific calculation method is as follows:

[0032]

[0033] ; in, Indicates the current state of the AGV. Lateral positional deviation; Indicates the current state of the AGV. Longitudinal positional deviation; Indicates the current state of the AGV. The deviation of the heading angle.

[0034] The target trajectory information is a pre-planned sequence of reference path points stored in the AGV control system.

[0035] S3.2: Obtain the acceleration response change characteristics and steering response change characteristics from the motion response change characteristics, perform state correlation analysis on the acceleration response change characteristics with the lateral position deviation and longitudinal position deviation, and determine the first influencing factor of the acceleration response deviation caused by load change on the position tracking performance; Specifically, the acceleration response change characteristics are obtained from the calculation results of step S2.1. This refers to the change in longitudinal acceleration caused by the change in load. respectively with lateral position deviation and longitudinal position deviation The state association analysis is performed, and its analysis logic is as follows: when When positive, it indicates that the AGV is currently in an unloading state, and its longitudinal acceleration is increased relative to the expected value. If the longitudinal position deviation is at this time... A positive increase indicates that the acceleration increment caused by load changes has already affected longitudinal position tracking; conversely, when... A negative value indicates that the AGV is currently under loading, and its longitudinal acceleration is lower than expected. Based on this, and considering the characteristics of acceleration response changes... The relationship between the rate of change of positional deviation and the first influencing factor is used to determine the first influencing factor. The calculation method is as follows: ; in, This represents the first influencing factor, used to quantify the degree of impact of acceleration response deviation caused by load changes on position tracking performance; and The pre-set correlation analysis coefficients were obtained from actual vehicle calibration. The sign function is used to determine the directional consistency between deviation and load change; the first term characterizes the impact of acceleration response change on longitudinal position tracking, and the second term characterizes the coupling effect of acceleration response change on lateral position tracking. First Influence Factor The sign of the symbol reflects the trend of how load changes affect position tracking, and its absolute value reflects the degree of influence.

[0036] S3.3: Perform state correlation analysis between the steering response change characteristics and the heading angle deviation to determine the second influencing factor of the steering response deviation caused by load change on the heading tracking performance; Specifically, the steering response change characteristics are obtained from the calculation results of step S2.1. This refers to the change in yaw acceleration caused by the change in load. deviation from heading angle The state association analysis is performed, and its analysis logic is as follows: when When positive, it indicates that the load change has caused the AGV's yaw acceleration to increase relative to the expectation. If the heading angle deviation is at this time... If the rate of change increases positively, it indicates that the steering response deviation caused by the load offset has already affected heading tracking. Based on this, and considering the characteristics of steering response change... The relationship between the second influencing factor and the rate of change of the heading angle deviation is used to determine the second influencing factor. The calculation method is as follows: ; in, This represents the second influencing factor, used to quantify the degree of impact of steering response deviation caused by load changes on heading tracking performance; The preset correlation analysis coefficients typically range from 0.5 to 2.0. They are related to the inherent parameters of the AGV, such as wheelbase, track width, vehicle weight, and tire lateral stiffness, and are obtained through actual vehicle calibration. The rate of change of the heading angle deviation is expressed as follows: ,in The sampling period; The sign function. Second influence factor. The sign of the symbol reflects the direction in which load changes affect heading tracking, and its absolute value reflects the degree of influence.

[0037] S3.4: Based on the first influence factor and the second influence factor, the trajectory tracking deviation data is corrected by load influence compensation to obtain effective tracking deviation data after motion response feature correction; Specifically, the lateral position deviation calculated in step S3.1 Longitudinal position deviation and heading angle deviation Subtract the deviation disturbance component caused by load variation from each component to obtain the corrected effective tracking deviation data. The effective tracking deviation data is calculated as follows: ; in, This indicates the corrected effective lateral position deviation; This indicates the corrected effective longitudinal position deviation; This indicates the corrected effective heading angle deviation; , and These are the preset compensation coefficients, where, The preferred value range is 0.05 to 0.5. The preferred value range is 0.05 to 0.5; The preferred value range for the factor is 0.02 to 0.2; the above compensation coefficient is used to adjust the impact factor. and The dimensions are unified to the deviation dimension, and its specific value is related to the kinematic parameters of the AGV and the controller structure, and is obtained through actual vehicle calibration. and The position deviation compensation coefficient is used to unify the dimensions of the first influencing factor to the dimensions of the lateral position deviation and the longitudinal position deviation, respectively. Its value is related to parameters such as the maximum acceleration and velocity loop gain of the AGV. This is the heading angle deviation compensation coefficient, used to unify the dimensions of the second influencing factor to the heading angle deviation dimension. Its value is related to parameters such as the AGV's wheelbase and maximum yaw rate. Typically... The value is less than and This is because the sensitivity of the angular deviation compensation coefficient is usually higher than that of the position deviation compensation coefficient. A smaller compensation coefficient can achieve effective heading deviation correction and avoid heading control oscillations caused by overcompensation.

[0038] Obtained from actual vehicle calibration, this data is used to unify the dimensions of influencing factors to the dimensions of deviation. Its value is related to the kinematic parameters of the AGV and the controller structure. The physical meaning of this correction process is that a portion of the deviation in the trajectory tracking deviation data is actually caused by changes in motion response due to load variations, rather than by insufficient tracking capability of the AGV control system. Therefore, it is necessary to remove this load-induced deviation component from the original deviation to obtain an effective deviation that truly reflects the tracking performance of the control system, thereby avoiding over-responding to load-induced deviations by the control system.

[0039] S3.5: Based on the effective tracking deviation data and combined with the real-time motion state information at the current moment, generate control adjustment requirement information characterizing the trajectory tracking performance under the current load condition. The control adjustment requirement information includes at least the position loop control adjustment requirement and the heading loop control adjustment requirement.

[0040] Specifically, the effective lateral position deviation calculated in step S3.4 Effective longitudinal position deviation and effective heading angle deviation Each with a preset position loop deviation threshold and heading loop deviation threshold The values ​​are compared, and corresponding control adjustment requirements are generated based on the magnitude and trend of the deviation. The preset position loop deviation threshold and heading loop deviation threshold are set as follows: Take a depth of 0.02m to 0.10m. The value is taken from 0.5° to 2.0°, and its specific value is determined based on the maximum allowable position tracking error and maximum heading tracking error of the actual application scenario of the AGV. A smaller value is used when the task requires high-precision tracking, and a larger value is used when the task allows for a larger tolerance. This value is configured by the operator during control system initialization according to task requirements. The calculation method for control adjustment requirements is as follows: ; in, This represents the position loop control adjustment requirement, used to characterize the control adjustment range needed for position tracking performance; This indicates the required adjustment amount for the heading loop control, used to characterize the control adjustment range needed for heading tracking performance; and The preset adjustment demand mapping function takes the effective tracking deviation and its rate of change as input and outputs the corresponding control adjustment demand. It is constructed using a piecewise linear function or a sigmoid function, with the effective tracking deviation and its rate of change as input variables. When the absolute value of the effective deviation is less than a preset deviation threshold, the mapping function outputs zero or a small adjustment demand. When the absolute value of the effective deviation exceeds the preset deviation threshold, the mapping function output monotonically increases with the deviation, and outputs a saturation value when the deviation reaches a preset saturation threshold. Simultaneously, the deviation rate of change term provides a damping effect, increasing the adjustment demand when the deviation tends to increase further and decreasing the adjustment demand when the deviation tends to decrease. and The value of reflects the additional adjustment required by the controller under the current load condition: when the effective deviation exceeds the preset threshold range, the control adjustment requirement increases, indicating that the control parameters need to be adjusted to a greater extent to suppress the tracking deviation caused by load changes; when the effective deviation is within the preset threshold range, the control adjustment requirement is small or zero, indicating that the current control parameters can still maintain sufficient tracking accuracy.

[0041] Step 4: Based on the control adjustment requirement information, adaptively adjust the trajectory tracking control parameters of the AGV to generate steering wheel control parameters that match the current load state, and then determine the steering control quantity and drive control quantity corresponding to the two steering wheels. In this embodiment, the logic for determining the steering wheel control parameters in step 4 is as follows: Based on the position loop control adjustment demand and the heading loop control adjustment demand in the control adjustment demand information, determine the position loop control parameter adjustment and the heading loop control parameter adjustment respectively. Among them, the adjustment amount of the position loop control parameters The calculation method is as follows Adjustment amount of heading loop control parameters The calculation method is as follows , and The preset position loop and heading loop adjustment step size coefficients are obtained from actual vehicle calibration and are used to map the control adjustment requirements to the actual parameter adjustment range. Their values ​​are related to the AGV's controller structure and response characteristics. The preferred value range is 0.01 to 0.1. The preferred value range is 0.005 to 0.05.

[0042] Then, based on the adjustment amount of the position loop control parameters, the proportional gain coefficient and integral gain coefficient of the position loop controller are adaptively adjusted to generate position loop control gain parameters that match the current load state; the position loop controller adopts a PI control structure, and its control law is... ,in This is the output of the position loop controller. For the effective tracking deviation of the position loop, This is the position loop proportional gain coefficient. This refers to the integral gain coefficient of the position loop controller. The proportional gain coefficient of the position loop controller is also relevant. and integral gain coefficient The adaptive adjustment method is as follows: ; in, This represents the adaptively adjusted position loop proportional gain coefficient. This represents the adaptively adjusted position loop integral gain coefficient; and These are the initial values ​​of the position loop proportional gain coefficient and the integral gain coefficient, respectively. They are preset parameters of the system and are obtained through offline tuning. and These are the preset proportional gain adjustment allocation coefficients and integral gain adjustment allocation coefficients, obtained from actual vehicle calibration, used to adjust the uniform position loop control parameters. It is allocated to two channels: proportional gain and integral gain. and The value of satisfies ,in The preferred value range is 0.6 to 0.8. The preferred value range is 0.2 to 0.4.

[0043] Simultaneously, based on the adjustment amount of the heading loop control parameters, the proportional gain coefficient and derivative gain coefficient of the heading loop controller are adaptively adjusted to generate heading loop control gain parameters that match the current load state; the heading loop controller adopts a PD control structure, and its control law is... ,in This is the output of the heading loop controller. For the effective tracking deviation of the heading loop, The rate of change of the heading angle deviation. This is the proportional gain coefficient of the heading loop. This represents the differential gain coefficient of the heading loop. The proportional gain coefficient of the heading loop controller. and differential gain coefficient The adaptive adjustment method is as follows: ; in, This represents the adaptively adjusted heading loop proportional gain coefficient; This represents the adaptively adjusted heading loop differential gain coefficient; and These are the initial values ​​of the yaw loop proportional gain coefficient and the differential gain coefficient, respectively. They are preset parameters of the system and are obtained through offline tuning. and These are the preset proportional gain adjustment allocation coefficients and derivative gain adjustment allocation coefficients, obtained from actual vehicle calibration, used to adjust the uniform heading loop control parameters. The gain is allocated to two channels: proportional gain and derivative gain. and The value of satisfies ,in The preferred value range is 0.5 to 0.7. The preferred value range is 0.3 to 0.5. During the adaptive adjustment process described above, it is necessary to ensure that the adjusted gain coefficient does not exceed the preset safety boundary. The upper and lower limits of each gain coefficient are predetermined by system stability analysis and stored in the AGV control system. After completing the adaptive adjustment, the position loop proportional gain coefficient, position loop integral gain coefficient, heading loop proportional gain coefficient, and heading loop differential gain coefficient together constitute the position loop control gain parameters and heading loop control gain parameters that match the current load state.

[0044] The position loop control gain parameters and the heading loop control gain parameters are input to the trajectory tracking controller. The trajectory tracking controller calculates the control quantity based on the real-time motion state information and effective tracking deviation data at the current moment, generating a vehicle control quantity that matches the current load state. The vehicle control quantity includes the vehicle target linear velocity and the vehicle target angular velocity. The calculation method for the vehicle control quantity is as follows: First, the position loop controller calculates the position loop output quantity based on the effective lateral position deviation and the effective longitudinal position deviation, and generates the vehicle target linear velocity command by combining it with a velocity feedforward term. The heading loop controller calculates the heading loop output quantity based on the effective heading angular deviation, and generates the vehicle target angular velocity command by combining it with a yaw rate feedforward term. The vehicle control quantity is expressed as: ; in, Indicates the target linear velocity of the entire vehicle; Indicates the target angular velocity of the entire vehicle; and These are the control input calculation functions for the position loop and the heading loop, respectively. Their specific forms are determined by the design structure of the trajectory tracking controller, and either PID control law or model predictive control law can be used. and Used for velocity feedforward compensation; Used for angular velocity feedforward compensation. The target linear velocity of the vehicle is the desired translational velocity of the AGV's center of mass in the vehicle coordinate system, and the target angular velocity of the vehicle is the desired rotational velocity of the AGV about the vertical axis of the center of mass. Together, they constitute the vehicle control quantity.

[0045] The steering wheel control parameters are characterized by vehicle control variables.

[0046] The target linear velocity and target angular velocity of the vehicle are input into the dual-steering wheel inverse kinematics module. Based on the geometric arrangement parameters of the dual steering wheels, inverse kinematics calculations are performed. The inverse kinematics formula for the dual-steering wheel omnidirectional AGV is: ; in, and These are the drive speed commands for the first and second steering wheels, respectively. and These are the steering angle commands for the first and second steering wheels, respectively. and These are the longitudinal distances from the first and second steering wheels to the center of mass of the AGV, respectively. and These are the lateral distances from the first and second steering wheels to the AGV's center of mass, respectively. These four geometric arrangement parameters are structural design parameters of the AGV and are pre-stored in the AGV control system during initialization. Through the aforementioned inverse kinematics calculation, the target linear velocity and target angular velocity of the entire vehicle are allocated to the two steering wheels, generating the steering control input corresponding to the first steering wheel. and drive control quantity and the steering control amount corresponding to the second steering wheel and drive control quantity .

[0047] Step 5: Perform motion control on the AGV according to the steering control quantity and drive control quantity, and acquire the vehicle status data after control in real time. Compare the vehicle status data with the target trajectory information and update the trajectory tracking deviation data to realize adaptive trajectory tracking control of the AGV under different load conditions. In this embodiment, the real-time acquisition of vehicle status data after control is specifically as follows: The steering control and drive control quantities corresponding to the first steering wheel are sent to the steering drive motor and travel drive motor of the first steering wheel, and the steering control and drive control quantities corresponding to the second steering wheel are sent to the steering drive motor and travel drive motor of the second steering wheel, driving the AGV to execute movement according to the motion command calculated in the current control cycle; specifically, firstly, the steering control quantity corresponding to the first steering wheel generated in step 4 is... and drive control quantity The steering drive motor and travel drive motor of the first steering wheel are sent out, and the steering control quantity corresponding to the second steering wheel is also sent out. and drive control quantity The steering drive motor and travel drive motor of the second steering wheel are issued commands. Each steering drive motor and travel drive motor of each steering wheel executes the corresponding action according to the received command signal, driving the AGV to move according to the motion command calculated in the current control cycle. During the issuance and execution of control commands, the AGV control system monitors the actual execution status of each motor in real time through closed-loop feedback, ensuring that the deviation between the actual steering angle and actual drive speed of the steering wheel and the command value remains within the preset execution accuracy range.

[0048] At the next sampling moment after the motion is performed, the updated vehicle status data is obtained in real time through the positioning detection module, the steering wheel status detection module and the load detection module. The updated vehicle status data includes updated vehicle pose data, updated dual steering wheel motion data and updated load status data.

[0049] Furthermore, the updated vehicle pose data is compared with the preset target trajectory information, and the trajectory tracking deviation data at the current moment is recalculated to obtain the updated trajectory tracking deviation data. Determine whether the updated trajectory tracking deviation data meets a preset trajectory tracking accuracy threshold condition; the trajectory tracking accuracy threshold condition is set as follows: the absolute value of the lateral position deviation is less than or equal to a preset lateral position deviation threshold. The absolute value of the longitudinal position deviation is less than or equal to the preset longitudinal position deviation threshold. Furthermore, the absolute value of the heading angle deviation is less than or equal to the preset heading angle deviation threshold. All three conditions mentioned above must be met simultaneously. The preset deviation threshold... The system presets parameters based on the actual application scenario and control precision requirements of the AGV. The preferred value range is 0.01m-0.05m. The preferred value range is 0.5°-2°, and the specific value is configured by the operator during the initialization of the control system according to the task requirements.

[0050] If the trajectory tracking accuracy threshold condition is met, the current control parameters remain unchanged, that is, the position loop control gain parameters and heading loop control gain parameters that have been adjusted and determined in step 4 are not readjusted, and the updated vehicle state data is used as the initial state data for the next control cycle. Then, step 1 is returned to begin the next control cycle. If the trajectory tracking accuracy threshold condition is not met, the updated trajectory tracking deviation data is used as the initial deviation data for the next control cycle, and the updated vehicle state data is used as the initial state data for the next control cycle. Then, return to step 2 and re-execute load change recognition, motion response feature extraction, adaptive adjustment of control parameters, and steering wheel control quantity calculation until the trajectory tracking deviation of the AGV under different load conditions meets the trajectory tracking accuracy threshold condition, thereby realizing the closed-loop adaptive trajectory tracking control of the AGV.

[0051] Table 1: Comparison of trajectory tracking deviations between traditional control and the method of this invention under different load capacities.

[0052] See also Figure 2 , Figure 3 To verify the effectiveness of the variable load adaptive trajectory tracking control method proposed in this invention, this embodiment conducted 20 sets of simulation experiments under progressively increasing load conditions ranging from 100kg to 1050kg. The maximum heading angle deviation and maximum lateral position deviation under the action of the traditional fixed parameter controller and the adaptive controller of this invention were recorded respectively. The experimental data show that as the load increases, the tracking deviation under both control methods shows an increasing trend, but the deviation value of the method of this invention under each load condition is significantly lower than that of the traditional control method. Regarding heading angle deviation, the traditional control method increased from 1.42° at 100kg to 3.55° at 1050kg, while the method of this invention only increased from 0.71° to 1.46°, representing an improvement in control accuracy of approximately 49.9% to 58.9%. Regarding lateral position deviation, the traditional control method increased from 31.6mm at 100kg to 84.5mm at 1050kg, while the method of this invention only increased from 18.2mm to 33.5mm, representing an improvement in control accuracy of approximately 42.4% to 60.3%. These data clearly demonstrate that the method proposed in this invention can effectively suppress the adverse effects of load changes on trajectory tracking accuracy under different load conditions.

[0053] Further analysis of the deviation variation characteristics in each load range reveals that in the low load range (100kg-400kg), the heading angle deviation growth rate of the method of this invention is approximately 0.00077° / kg, while that of the traditional method is approximately 0.0021° / kg. This indicates that the method of this invention can effectively maintain heading stability under light load conditions. In the heavy load range (700kg-1050kg), the heading angle deviation of the traditional control method increases sharply, with a maximum growth rate of 0.0047° / kg, while the deviation growth rate of the method of this invention remains at approximately 0.0013° / kg. This demonstrates that the designed adaptive gain adjustment mechanism can effectively compensate for the increased system inertia and response lag caused by the increased mass under heavy load conditions, significantly improving the trajectory tracking robustness of the AGV under heavy load and variable mass conditions. This verifies the effectiveness and engineering practical value of the method of this invention across the entire load range.

[0054] The above formulas are all dimensionless calculations. The formulas are derived from software simulations based on a large amount of collected data to obtain the most recent real-world results. The preset parameters in the formulas are set by those skilled in the art according to the actual situation.

[0055] The above embodiments can be implemented, in whole or in part, by software, hardware, firmware, or any other combination thereof. When implemented in software, the above embodiments can be implemented, in whole or in part, as a computer program product. Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented by electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution.

[0056] 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; they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment, depending on actual needs.

[0057] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any changes or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application.

Claims

1. A variable load adaptive trajectory tracking control method for a dual-steering wheel omnidirectional AGV, characterized in that, Specifically, it includes: Step 1: Acquire vehicle status data and load status data during AGV operation. The vehicle status data includes vehicle pose data, dual steering wheel angle data, and drive wheel motion data. The load status data includes load mass and load change. Based on the vehicle status data and load status data, construct status feature data corresponding to the current operating condition. Step 2: Based on the state feature data, identify the current load change of the AGV, determine the motion response change characteristics caused by the load change, and update the pre-built vehicle motion state model based on the motion response change characteristics to obtain real-time motion state information that matches the current load state. Step 3: Based on the real-time motion state information and the preset target trajectory information, determine the trajectory tracking deviation data of the AGV, and combine the motion response change characteristics to perform state correlation analysis on the trajectory tracking deviation data to obtain control adjustment requirements information characterizing the trajectory tracking performance under the current load condition; Step 4: Based on the control adjustment requirement information, adaptively adjust the trajectory tracking control parameters of the AGV to generate steering wheel control parameters that match the current load state, and then determine the steering control quantity and drive control quantity corresponding to the two steering wheels. Step 5: Perform motion control on the AGV based on the steering control and drive control quantities, and acquire the vehicle status data after control in real time. Compare the vehicle status data with the target trajectory information and update the trajectory tracking deviation data to achieve adaptive trajectory tracking control of the AGV under different load conditions.

2. The variable load adaptive trajectory tracking control method for a dual-steering wheel omnidirectional AGV according to claim 1, characterized in that: Step 1 specifically includes: By using a positioning detection module, a steering wheel status detection module, and a load detection module installed on the AGV body, vehicle pose data, dual steering wheel motion data, and load status data during AGV operation are obtained, respectively. Among them, the positioning detection module obtains the AGV's current position coordinates and vehicle heading angle in the global coordinate system to form vehicle pose data. The vehicle pose data is composed of the AGV's lateral position coordinates, longitudinal position coordinates, and vehicle heading angle at the current moment. The steering wheel status detection module collects the steering angle and driving speed of the two steering wheels respectively to form dual steering wheel motion data. The dual steering wheel motion data is composed of the steering angle of the first steering wheel at the current moment, the steering angle of the second steering wheel at the current moment, the driving speed of the first steering wheel at the current moment, and the driving speed of the second steering wheel at the current moment.

3. The variable load adaptive trajectory tracking control method for a dual-steering wheel omnidirectional AGV according to claim 2, characterized in that: The load detection module obtains the current load status information of the AGV and generates load status data based on the characteristic parameters that affect the vehicle's motion performance. The load status data is composed of the AGV's load mass at the current moment and the current load change. The current load change is used to characterize the load change between adjacent sampling moments, and it is determined by the difference between the load mass obtained at the current moment and the load mass obtained at the previous moment. The vehicle pose data, dual steering wheel motion data, and load status data are processed for time synchronization, and the status feature data corresponding to the current operating condition is constructed according to a preset arrangement rule.

4. The variable load adaptive trajectory tracking control method for a dual-steering wheel omnidirectional AGV according to claim 1, characterized in that: Step 2 specifically includes: S2.1: Based on the load status data in the state feature data, identify the current load change of the AGV. The load change includes the direction and rate of change of the load mass. Based on the direction and rate of change of the load mass, determine the vehicle motion response change characteristics caused by the load change. The motion response change characteristics include acceleration response change characteristics and steering response change characteristics. S2.2: Based on the acceleration response change characteristics, update the velocity state equation in the pre-constructed vehicle motion state model to obtain a velocity state update equation that matches the current load state; S2.3: Based on the steering response change characteristics, update the attitude state equation in the pre-constructed vehicle motion state model to obtain an attitude state update equation that matches the current load state. S2.4: Based on the velocity state update equation and attitude state update equation, the vehicle pose data and dual steering wheel motion data in the state feature data at the current moment are used as inputs to the updated vehicle motion state model to calculate real-time motion state information that matches the current load state. The real-time motion status information includes at least the predicted lateral speed, predicted longitudinal speed, and predicted yaw rate of the AGV under the current load condition.

5. The variable load adaptive trajectory tracking control method for a dual-steering wheel omnidirectional AGV according to claim 4, characterized in that: Step 3 specifically includes: S3.1: Based on the real-time motion status information and the preset target trajectory information, determine the trajectory tracking deviation data of the AGV in the global coordinate system. The trajectory tracking deviation data includes lateral position deviation, longitudinal position deviation and heading angle deviation. S3.2: Obtain the acceleration response change characteristics and steering response change characteristics from the motion response change characteristics, perform state correlation analysis on the acceleration response change characteristics with the lateral position deviation and longitudinal position deviation, and determine the first influencing factor of the acceleration response deviation caused by load change on the position tracking performance; S3.3: Perform state correlation analysis between the steering response change characteristics and the heading angle deviation to determine the second influencing factor of the steering response deviation caused by load change on the heading tracking performance; S3.4: Based on the first influence factor and the second influence factor, the trajectory tracking deviation data is corrected by load influence compensation to obtain effective tracking deviation data after motion response feature correction; S3.5: Based on the effective tracking deviation data and combined with the real-time motion state information at the current moment, generate control adjustment requirement information characterizing the trajectory tracking performance under the current load condition. The control adjustment requirement information includes at least the position loop control adjustment requirement and the heading loop control adjustment requirement.

6. The variable load adaptive trajectory tracking control method for a dual-steering wheel omnidirectional AGV according to claim 5, characterized in that: The logic for determining the steering wheel control parameters in step 4 is as follows: Based on the position loop control adjustment demand and the heading loop control adjustment demand in the control adjustment demand information, determine the position loop control parameter adjustment and the heading loop control parameter adjustment respectively. Based on the adjustment amount of the position loop control parameters, the proportional gain coefficient and integral gain coefficient of the position loop controller are adaptively adjusted to generate position loop control gain parameters that match the current load state. Based on the adjustment amount of the heading loop control parameters, the proportional gain coefficient and derivative gain coefficient of the heading loop controller are adaptively adjusted to generate heading loop control gain parameters that match the current load state. The position loop control gain parameter and the heading loop control gain parameter are input to the trajectory tracking controller. The trajectory tracking controller calculates the control quantity based on the real-time motion state information and effective tracking deviation data at the current moment, and generates a vehicle control quantity that matches the current load state. The vehicle control quantity includes the vehicle target linear velocity and the vehicle target angular velocity. The steering wheel control parameters are characterized by vehicle control variables.

7. The variable load adaptive trajectory tracking control method for a dual-steering wheel omnidirectional AGV according to claim 6, characterized in that: The logic for determining the steering control quantity and the drive control quantity is as follows: The target linear velocity and target angular velocity of the vehicle are input into the dual-steering wheel inverse kinematics module. Based on the geometric arrangement parameters of the dual steering wheels, the inverse kinematics calculation is performed, and the target linear velocity and target angular velocity of the vehicle are distributed to the two steering wheels, generating the steering control and drive control quantities corresponding to the first steering wheel and the steering control and drive control quantities corresponding to the second steering wheel, respectively.

8. The variable load adaptive trajectory tracking control method for a dual-steering wheel omnidirectional AGV according to claim 7, characterized in that: Real-time acquisition of vehicle status data after control, specifically: The steering control and drive control quantities corresponding to the first steering wheel are sent to the steering drive motor and the travel drive motor of the first steering wheel, and the steering control and drive control quantities corresponding to the second steering wheel are sent to the steering drive motor and the travel drive motor of the second steering wheel, driving the AGV to execute motion according to the motion command calculated in the current control cycle. At the next sampling moment after the motion is performed, the updated vehicle status data is obtained in real time through the positioning detection module, the steering wheel status detection module and the load detection module. The updated vehicle status data includes updated vehicle pose data, updated dual steering wheel motion data and updated load status data.

9. A variable load adaptive trajectory tracking control method for a dual-steering wheel omnidirectional AGV according to claim 8, characterized in that: The updated vehicle pose data is compared with the preset target trajectory information, and the trajectory tracking deviation data at the current moment is recalculated to obtain the updated trajectory tracking deviation data. Determine whether the updated trajectory tracking deviation data meets the preset trajectory tracking accuracy threshold condition; If the trajectory tracking accuracy threshold condition is met, the current control parameters remain unchanged, and the updated vehicle status data is used as the initial status data for the next control cycle, returning to step 1; If the trajectory tracking accuracy threshold condition is not met, the updated trajectory tracking deviation data is used as the initial deviation data for the next control cycle, and the updated vehicle state data is used as the initial state data for the next control cycle. Then, return to step 2 and re-execute load change recognition, motion response feature extraction, adaptive adjustment of control parameters, and steering wheel control quantity calculation until the trajectory tracking deviation of the AGV under different load conditions meets the trajectory tracking accuracy threshold condition, thereby realizing the closed-loop adaptive trajectory tracking control of the AGV.

Images