Method, device and system for controlling travel of a vibratory roller

Abstract

This invention relates to the field of unmanned vibratory roller technology, specifically disclosing an unmanned vibratory roller driving control method, device, and system, comprising: acquiring the positioning information of the unmanned vibratory roller; processing the positioning information using an unscented Kalman filter to obtain a real-time state estimate of the unmanned vibratory roller; acquiring the trajectory planning result of the unmanned vibratory roller, and calculating the error between the trajectory planning result and the real-time state estimate to obtain an error result; inputting the error result to an LQR controller to obtain unmanned vibratory roller control parameters; and controlling the driving of the unmanned vibratory roller according to the unmanned vibratory roller control parameters. The unmanned vibratory roller driving control method provided by this invention can improve the robustness of unmanned vibratory roller control and the accuracy of trajectory tracking in complex environments.

Description

Technical Field

[0001] This invention relates to the field of unmanned vibratory roller technology, and in particular to an unmanned vibratory roller driving control method, an unmanned vibratory roller driving control device, and an unmanned vibratory roller driving control system. Background Technology

[0002] With the development of modern industrialization, vibratory rollers, due to their high production efficiency and excellent performance, have been widely used in infrastructure construction such as roads, airports, ports, and dams. With the rapid advancement of big data and artificial intelligence technologies, unmanned vibratory roller technology faces new challenges, requiring higher efficiency, stronger safety, and greater reliability.

[0003] Accurate tracking of the driving trajectory is one of the key issues in the unmanned driving control technology of vibratory rollers. Trajectory tracking involves designing the horizontal motion of the unmanned vibratory roller, including its longitudinal and lateral movements. Currently, the traditional PID control (proportional-integral-derivative) method is widely used due to its simplicity, ease of implementation, and wide applicability. However, due to factors such as the high-frequency vibration disturbances of the vibratory roller and the complex types of construction road surfaces, the PID control method's sensitivity to disturbances becomes particularly prominent.

[0004] Existing technologies include several unmanned vibratory roller driving control methods, but all suffer from insurmountable drawbacks. For example, one implementation involves calculating the actual tracking distance error of the roller based on its actual position and target trajectory. Next, a disturbance observer is used to obtain the total distance error disturbance, and an outer-loop control is constructed to calculate the target heading angle. Then, the target steering wheel angle is calculated using the target heading angle, the total heading disturbance obtained from the instantaneous disturbance observer, and the heading angle disturbance obtained from the steering system model parameter learner. Finally, a steering wheel angle controller controls the steering wheel to bring the actual heading angle of the roller closer to the target value. This technology, due to the addition of the disturbance observer, requires more advanced hardware and software, increasing costs and making the control system more complex, requiring more frequent maintenance and more complex fault handling procedures. Another approach to existing technologies involves updating the occupancy grid map using sensors such as LiDAR, cameras, and millimeter-wave radar to generate collision-free constraints with the static environment. Based on the identified types of dynamic obstacles, their positions and contours are determined, and corresponding collision-free constraints are established. Secondly, based on the kinematic model of the roller, the control sequence in the predicted time domain is solved online, guiding the execution of the underlying mechanism to achieve precise control of the unmanned roller. However, this model prediction contour control method does not consider the impact of noise disturbances under complex working conditions, which will challenge the robustness of the controller in actual operating conditions. Furthermore, this method requires completing model prediction and optimization calculations within each control cycle, placing high demands on computational and system real-time performance.

[0005] In addition, the working surface of the vibratory roller may be very rough, including steep slopes, soft ground or unstable soil and rocks, which poses a challenge to the robustness of the controller.

[0006] Therefore, how to provide a method that can improve the robustness of unmanned vibratory roller control and the accuracy of trajectory tracking in complex environments has become a technical problem that urgently needs to be solved by those skilled in the art. Summary of the Invention

[0007] This invention provides an unmanned vibratory roller driving control method, an unmanned vibratory roller driving control device, and an unmanned vibratory roller driving control system, which solves the problems in related technologies that cannot improve the robustness of unmanned vibratory roller control and the accuracy of trajectory tracking in complex environments.

[0008] As a first aspect of the present invention, a method for controlling the movement of an unmanned vibratory roller is provided, comprising:

[0009] The positioning information of the unmanned vibratory roller is obtained, and the positioning information includes at least the vehicle body coordinates, heading angle and articulation point steering angle;

[0010] The positioning information is processed by an unscented Kalman filter to obtain a real-time state estimate of the unmanned vibratory roller.

[0011] Obtain the trajectory planning results of the unmanned vibratory roller, and calculate the error between the trajectory planning results and the real-time state estimate to obtain the error result;

[0012] The error results are input into the LQR controller to obtain the control parameters for the unmanned vibratory roller.

[0013] The movement of the unmanned vibratory roller is controlled according to the control parameters of the unmanned vibratory roller.

[0014] Furthermore, the positioning information is processed using an unscented Kalman filter, including:

[0015] Construct an unscented Kalman filter;

[0016] The positioning information is input into the unscented Kalman filter for processing to obtain the vehicle body coordinate estimate.

[0017] Based on the vehicle body size information, the estimated vehicle body coordinates are transformed into coordinate points to obtain the real-time position estimate of the roller.

[0018] Furthermore, an unscented Kalman filter is constructed, including:

[0019] The motion of the unmanned vibratory roller is simplified to planar motion, and the state-space equation of the vibratory roller-vehicle body is established based on the kinematic model.

[0020] The state transition function is determined based on the vibration roller-vehicle state space equation.

[0021] The sigma points are predicted using the state transition function to obtain the Kalman gain and covariance matrix, thus completing the construction of the unscented Kalman filter.

[0022] Furthermore, the expression for the state-space equation of the vibration roller-vehicle body is as follows:

[0023] ,

[0024] in, Indicates the position and speed of the vehicle's center of gravity. This represents the velocity component of the vehicle body in the X direction in the global coordinate system. This represents the velocity component of the vehicle body in the Y direction in the global coordinate system. This represents the angular velocity of the vehicle's center of mass in the global coordinate system. This indicates the angle difference between the car body and the roller. This represents the distance from the center of mass of the roller to the hinge point. This indicates the distance from the vehicle's center of gravity to the hinge point;

[0025] The expression for the state transition function is:

[0026] ,

[0027] in, State variables , Indicates time, Indicates the sampling time of the discrete system; This represents the X-axis coordinate of the vehicle's center of mass in the global coordinate system. This represents the Y-axis coordinate of the vehicle's center of mass in the global coordinate system. The heading angle represents the center of gravity of the vehicle body.

[0028] Further, the sigma point is predicted according to the state transition function to obtain the Kalman gain and covariance matrix, including:

[0029] Initialize to generate sigma points and determine the state estimation vector and covariance matrix at the initialization time;

[0030] The sigma point is nonlinearly transformed through the state transition function to obtain the predicted state estimate of the sigma point;

[0031] Calculate the mean and covariance of the predicted state based on the state estimates of the predicted sigma points;

[0032] The Kalman gain is calculated based on the mean and covariance of the predicted state, and the state estimate and covariance matrix are updated.

[0033] Furthermore, the error result is input to the LQR controller to obtain the control parameters for the unmanned vibratory roller, including:

[0034] Construct an LQR controller to determine the control variables;

[0035] The error result is controlled according to the control quantity to obtain the control parameters for the unmanned vibratory roller.

[0036] Furthermore, an LQR controller is constructed to determine the control variables, including:

[0037] Construct error-based state-space equations;

[0038] Determine the cost function and the output weight matrix;

[0039] Design an LQR controller based on the error-based state-space equation, cost function, and output weight matrix;

[0040] The control quantity is determined based on the LQR controller.

[0041] Further, the cost function and output weight matrix are determined, including:

[0042] The error-based state-space equation is discretized to obtain a linear time-varying prediction model for the vibrating mill-drum.

[0043] The LQR cost calculation equation is determined based on the linear time-varying prediction model of the vibratory roller.

[0044] The output weight matrix is ​​determined based on the LQR cost calculation equation.

[0045] As another aspect of the present invention, an unmanned vibratory roller driving control device is provided for implementing the unmanned vibratory roller driving control method described above, wherein the device includes:

[0046] The positioning information acquisition module is used to acquire the positioning information of the unmanned vibratory roller, and the positioning information includes at least the vehicle body coordinates, heading angle and articulation point steering angle;

[0047] The Kalman filter module is used to process the positioning information through an unscented Kalman filter to obtain the real-time state estimate of the unmanned vibratory roller.

[0048] The error calculation module is used to obtain the trajectory planning results of the unmanned vibratory roller, and to perform error calculation based on the trajectory planning results and the real-time state estimate to obtain the error result;

[0049] The LQR control module is used to input the error results to the LQR controller to obtain the control parameters of the unmanned vibratory roller.

[0050] The driving control module is used to control the driving of the unmanned vibratory roller according to the control parameters of the unmanned vibratory roller.

[0051] As another aspect of the present invention, an unmanned vibratory roller driving control system is provided, comprising: a positioning device, a decision-making device, an execution device, and the unmanned vibratory roller driving control device described above, wherein the positioning device, the decision-making device, and the execution device are all communicatively connected to the unmanned vibratory roller driving control device.

[0052] The positioning device is used to collect positioning information of the unmanned vibratory roller;

[0053] The decision-making device is used to plan the trajectory of the unmanned vibratory roller and obtain the trajectory planning result.

[0054] The unmanned vibratory roller driving control device is used to process the positioning information according to the unscented Kalman filter, calculate the error according to the trajectory planning result and the real-time state estimate, and input the error result to the LQR controller to obtain the unmanned vibratory roller control parameters.

[0055] The actuator is used to control the movement of the unmanned vibratory roller according to the control parameters of the unmanned vibratory roller.

[0056] The unmanned vibratory roller driving control method provided by this invention processes the acquired positioning information using an unscented Kalman filter, calculates the error between the planning result and the real-time state estimate output by the unscented Kalman filter, and inputs the error result to an LQR controller to obtain the unmanned vibratory roller control parameters. Based on these control parameters, the unmanned vibratory roller's movement is controlled. This unmanned vibratory roller driving control method combines an unscented Kalman filter with an LQR controller, enabling the unmanned vibratory roller to obtain relatively accurate control error values ​​even when the system measurement parameters contain disturbances, and achieving accurate tracking of the desired trajectory according to the proposed control priority. Therefore, the unmanned vibratory roller driving control method of this invention can improve the robustness of unmanned vibratory roller control and the accuracy of trajectory tracking in complex environments. Attached Figure Description

[0057] The accompanying drawings are provided to further illustrate the invention and form part of the specification. They are used together with the following detailed description to explain the invention, but do not constitute a limitation thereof.

[0058] Figure 1 The flowchart shows the unmanned vibratory roller driving control method provided by the present invention.

[0059] Figure 2 This is a flowchart illustrating how the unscented Kalman filter, provided by the present invention, processes positioning information.

[0060] Figure 3 The flowchart for constructing an unscented Kalman filter provided by this invention.

[0061] Figure 4 The flowchart for obtaining control parameters of an unmanned vibratory roller provided by the present invention.

[0062] Figure 5 A schematic diagram of the data flow for the unmanned vibratory roller driving control method provided by the present invention.

[0063] Figure 6 The structural block diagram of the unmanned vibratory roller driving control device provided by the present invention.

[0064] Figure 7 This is a structural block diagram of the unmanned vibratory roller driving control system provided by the present invention.

[0065] Figure 8 The structural block diagram shows a specific implementation of the unmanned vibratory roller driving control system provided by the present invention. Detailed Implementation

[0066] It should be noted that, unless otherwise specified, the embodiments and features described in the present invention can be combined with each other. The present invention will now be described in detail with reference to the accompanying drawings and embodiments.

[0067] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.

[0068] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate for the embodiments of the invention described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0069] Currently, the operating surfaces of unmanned vibratory rollers may be very rugged, including steep slopes, soft ground, or unstable soil and rocks, which poses a challenge to the robustness of the controller. To address the shortcomings and problems of traditional unmanned vibratory roller control methods, such as poor robustness in complex environments and difficulty in system parameter tuning leading to unsatisfactory control performance, this embodiment provides an unmanned vibratory roller travel control method. Figure 1 This is a flowchart of the unmanned vibratory roller driving control method provided by an embodiment of the present invention, such as... Figure 1 As shown, it includes:

[0070] S100. Obtain the positioning information of the unmanned vibratory roller, wherein the positioning information includes at least the vehicle body coordinates, heading angle and articulation point steering angle;

[0071] In this embodiment of the invention, the positioning device is used to collect the positioning information of the unmanned vibratory roller in real time. The positioning information includes at least the vehicle body coordinates, heading angle and articulation point steering angle.

[0072] Specifically, the positioning device is capable of determining the precise position and heading of the unmanned vibratory roller, and may include a GPS global positioning system and an inertial measurement unit (IMU), wherein the GPS receiver and the IMU may be installed inside the cab of the unmanned vibratory roller, and the positioning antenna is located at the top of the cab.

[0073] S200. The positioning information is processed by an unscented Kalman filter to obtain the real-time state estimate of the unmanned vibratory roller.

[0074] In this embodiment of the invention, the positioning information is processed by an unscented Kalman filter to obtain the vehicle body coordinate estimate. Since the vehicle body size is fixed, the coordinate point can be transformed based on the vehicle body coordinate estimate to obtain the position information of the roller at the current moment.

[0075] S300: Obtain the trajectory planning result of the unmanned vibratory roller, and calculate the error based on the trajectory planning result and the real-time state estimate to obtain the error result;

[0076] In this embodiment of the invention, the decision-making device can specifically output trajectory planning results. Specifically, the decision-making device can generate trajectory planning results for the unmanned vibratory roller based on real-time sensing information for the controller in the system. The specific implementation of trajectory planning is well known to those skilled in the art and will not be described in detail here. Error calculation is performed based on the trajectory planning results and the real-time state estimate to obtain the error result.

[0077] S400. Input the error result into the LQR controller to obtain the control parameters for the unmanned vibratory roller.

[0078] In this embodiment of the invention, the error result is input to the LQR controller, which can process the error result to obtain the control parameters of the unmanned vibratory roller.

[0079] S500: Control the movement of the unmanned vibratory roller according to the control parameters of the unmanned vibratory roller.

[0080] Specifically, the unmanned vibratory roller is controlled according to the aforementioned unmanned vibratory roller control parameters to complete the unmanned vibratory roller driving control process.

[0081] In summary, the unmanned vibratory roller driving control method provided by this invention processes the acquired positioning information using an unscented Kalman filter, calculates the error based on the planning results and the real-time state estimate output by the unscented Kalman filter, and inputs the error result to the LQR controller to obtain the unmanned vibratory roller control parameters. Based on these control parameters, the unmanned vibratory roller's movement is controlled. This unmanned vibratory roller driving control method combines an unscented Kalman filter with an LQR controller, enabling the unmanned vibratory roller to obtain relatively accurate control error values ​​even when the system measurement parameters contain disturbances, and achieving accurate tracking of the desired trajectory according to the proposed control priority. Therefore, the unmanned vibratory roller driving control method of this invention can improve the robustness of unmanned vibratory roller control and the accuracy of trajectory tracking in complex environments.

[0082] In this embodiment of the invention, the positioning information is processed using an unscented Kalman filter, such as... Figure 2 As shown, it includes:

[0083] S210, Construct an unscented Kalman filter;

[0084] In this embodiment of the invention, the location information is processed based on the unscented Kalman filter. First, an unscented Kalman filter needs to be constructed according to the target.

[0085] Specifically, an unscented Kalman filter is constructed, such as... Figure 3 As shown, it includes:

[0086] S211. Simplify the motion of the unmanned vibratory roller into planar motion, and establish the state-space equation of the vibratory roller-vehicle body based on the kinematic model;

[0087] The expression for the state-space equation of the vibratory roller-vehicle body is:

[0088] ,

[0089] in, Indicates the position and speed of the vehicle's center of gravity. This represents the velocity component of the vehicle body in the X direction in the global coordinate system. This represents the velocity component of the vehicle body in the Y direction in the global coordinate system. This represents the angular velocity of the vehicle's center of mass in the global coordinate system. This indicates the angle difference between the car body and the roller. This represents the distance from the center of mass of the roller to the hinge point. This indicates the distance from the vehicle's center of gravity to the hinge point.

[0090] S212. Determine the state transition function based on the vibration roller-vehicle state space equation;

[0091] In this embodiment of the invention, the expression of the state-space equation of the vibration roller-vehicle body is discretized to obtain the state transition function. .

[0092] In this embodiment of the invention, the expression of the state transition function is:

[0093] ,

[0094] in, State variables , Indicates time, Indicates the sampling time of the discrete system; This represents the X-axis coordinate of the vehicle's center of mass in the global coordinate system. This represents the Y-axis coordinate of the vehicle's center of mass in the global coordinate system. The heading angle represents the center of gravity of the vehicle body.

[0095] S213. Based on the state transition function, the sigma point is predicted to obtain the Kalman gain and covariance matrix, thereby completing the construction of the unscented Kalman filter.

[0096] In this embodiment of the invention, the sigma points are nonlinearly transformed based on the state transition function, and then the obtained predicted sigma points are processed to complete the construction of the unscented Kalman filter.

[0097] Specifically, the sigma point is predicted based on the state transition function to obtain the Kalman gain and covariance matrix, including:

[0098] 1) Initialize to generate sigma points and determine the state estimation vector and covariance matrix at the initialization time;

[0099] initialization, This is the initial time. After initialization, sigma points are generated, and these points are evenly distributed with respect to the estimated state vector. The left and right ends (a total of 2n+1) are related to the state estimation vector. Covariance Matrix have:

[0100] , for ,

[0101] , for ,

[0102] in, This represents the scaling parameter. , Represents the state dimension. express The state estimate at time t is the central sigma point. Indicates the first Sigma points.

[0103] 2) The sigma point is subjected to a nonlinear transformation through the state transition function to obtain the predicted state estimate of the sigma point;

[0104] In this embodiment of the invention, the expression for the state estimate of the predicted sigma point is:

[0105] .

[0106] 3) Calculate the mean and covariance of the predicted state based on the state estimates of the predicted sigma points;

[0107] In this embodiment of the invention, the specific calculation process is as follows:

[0108] ,

[0109] ,

[0110] in, and All represent the prediction weights of the sigma points. Represents the process noise covariance matrix;

[0111] ,

[0112] ,

[0113] ,

[0114] in, Used to control the distribution of sigma points This represents a parameter related to the prior distribution information of the state, which is used to adjust... This can improve the accuracy of the covariance matrix. For a Gaussian distribution, This is optimal; m and c are used to distinguish the categories of weight coefficients.

[0115] 4) Calculate the Kalman gain based on the mean and covariance of the predicted state and update the state estimate and covariance matrix.

[0116] In this embodiment of the invention, the predicted sigma point is obtained by taking one step forward using the following formula to obtain the predicted k-time measured sigma point at time k-1. :

[0117] ;

[0118] Combine the predicted sigma point at time k with the prediction at time k-1. The expression is given, and the mean and covariance of the predicted sigma point at time k are calculated one step ahead from the predicted sigma point:

[0119] ,

[0120] ,

[0121] ,

[0122] in, and Let these represent the measurement covariance matrix and the state-measurement cross-covariance matrix, respectively. This represents the measurement noise covariance matrix.

[0123] Finally, the Kalman gain is calculated and the state estimate and covariance matrix are updated.

[0124] ,

[0125] ,

[0126] ,

[0127] in, This represents the Kalman gain at time k after the update. Indicates after the update Time-varying covariance matrix express The time-state estimation error covariance matrix, express The actual value measured by the time sensor. express State estimate at time step.

[0128] S220. Input the positioning information into the unscented Kalman filter for processing to obtain the vehicle body coordinate estimation value;

[0129] In this embodiment of the invention, the above-mentioned positioning information can be input into an unscented Kalman filter to obtain the vehicle body coordinate estimate.

[0130] S230. Based on the vehicle body size information, the estimated vehicle body coordinates are converted into coordinate points to obtain the estimated real-time position of the roller.

[0131] Since the vehicle body dimensions are fixed, the estimated vehicle body coordinates can be transformed into coordinate points based on the estimated vehicle body coordinates to obtain the real-time position estimate of the rollers.

[0132] In this embodiment of the invention, the error result is input to the LQR controller to obtain the control parameters for the unmanned vibratory roller, such as... Figure 4 As shown, it includes:

[0133] S410. Construct an LQR controller to determine the control variables;

[0134] In this embodiment of the invention, an error-based state-space equation can be constructed based on the structure of the unmanned vibratory roller, thereby determining the control quantity of the LQR controller.

[0135] Specifically, an LQR controller is constructed to determine the control variables, including:

[0136] (1) Construct the state-space equation based on the error;

[0137] Considering the longitudinal symmetry of the unmanned vibratory roller structure, and that the tires and steel wheels are in point contact with the ground, and simplifying the motion of the unmanned vibratory roller to planar motion, the velocity of its roller's center of mass is... Decompose along the X and Y axes:

[0138] ,

[0139] in, and These represent the velocity at the center of mass in the roller coordinate system and the angle between it and the global coordinate system X, respectively.

[0140] Assuming there is no lateral or longitudinal slippage at the hinge point of the vibratory roller, the vector sum of the vehicle body velocities at that hinge point should be equal to the vector sum of the roller velocities at that hinge point:

[0141] ,

[0142] in, and These represent the distances from the center of mass of the roller and the center of mass of the vehicle body to the hinge point, respectively. and These represent the angular velocities of the roller and the vehicle body, respectively.

[0143] velocity vector and Switch to the coordinate axis where the drum is located to obtain:

[0144] ,

[0145] ,

[0146] in, This indicates the angle difference between the vehicle body and the roller.

[0147] according to and Combining the two expressions, we can obtain the differential equation of angular velocity in the drum coordinate system:

[0148] ,

[0149] This The expression for the velocity of the roller's center of mass By simultaneously establishing the expressions for the X and Y axes, the state-space equations of the vibratory roller-drum are constructed:

[0150] ,

[0151] Since the above vibratory roller-drum state-space equations are a nonlinear system of equations, to facilitate the solution by the LQR controller, they are linearized through Taylor expansion to obtain the error-based state control equations:

[0152] ,

[0153] in, , , , , ,

[0154] in, This represents the state matrix, used to describe the relationships between system state variables; Represents state variables; This represents the input matrix, used to describe the influence of the input on the states; Indicates system input; Indicates the roller speed; Indicates the hinge point angle; Indicates the desired speed of the roller; This indicates the distance from the center of mass of the roller to the hinge point; This indicates the distance from the vehicle's center of gravity to the hinge point; , and Indicate the desired point coordinates and desired heading angle; Indicates the expected curvature of the track point;

[0155] Discretizing the error-based state control equations yields a linear time-varying prediction model for the vibratory roller-drum system.

[0156] ,

[0157] in, , , , ,

[0158] Where T represents the offline system sampling time.

[0159] (2) Determine the cost function and the output weight matrix;

[0160] In this embodiment of the invention, the cost function and the output weight matrix are determined according to the error-based state-space equation described above.

[0161] Specifically, determining the cost function and the output weight matrix includes:

[0162] (21) Discretize the error-based state-space equation to obtain a linear time-varying prediction model for the vibrating roller.

[0163] (22) Determine the LQR cost calculation equation based on the linear time-varying prediction model of the vibratory roller-drum;

[0164] In this embodiment of the invention, the LQR cost calculation equation is expressed as follows:

[0165] ,

[0166] Where n=2, , , This represents the control effect matrix, used to indicate the degree of importance the controller places on state variables. This represents the output weight matrix, used to measure the impact of different outputs on system performance, with weighting coefficients. , and These are used to adjust the controller's sensitivity to longitudinal, lateral, and heading errors, respectively. and Adjust the speed control and steering angle control values ​​separately. If the lateral deviation rate of change or the curvature of the reference path... When the reference path is large, a smaller weight can be selected to improve following performance and reduce the lateral deviation between the actual driving path and the reference path. , and corner weight This makes the controller more sensitive to changes in steering angle, reduces the restriction on the front wheel steering angle, and ensures the tracking accuracy of the vibratory roller; if the rate of change of lateral deviation or the curvature of the reference path... If the value is small, a larger weight can be selected. , and corner weight This suppresses steering angle sway. The weight of longitudinal control... , Similarly.

[0167] By designing an output weight matrix, the impact of different outputs on system performance can be measured, thereby further improving the overall system performance and enhancing its robustness.

[0168] (23) Determine the output weight matrix according to the LQR cost calculation equation.

[0169] Based on the cost calculation equation and output weight matrix designed in the above steps, an LQR controller is designed, and the horizontal and vertical control feedback quantities are calculated to achieve dynamic control of the system.

[0170] Substitute into the solution to the discrete-time Riccati equation, and when The matrix can be obtained in time. ,in This is the state feedback gain matrix.

[0171] (3) Design an LQR controller based on the error-based state-space equation, cost function, and output weight matrix;

[0172] (4) Determine the control quantity based on the LQR controller.

[0173] Based on the state feedback gain matrix obtained above, the controller gain matrix coefficients are calculated. Finally, based on the controller gain matrix... Error information with current state Calculate the horizontal and vertical control values. Adding the expected value yields the actual control quantity. :

[0174] ,

[0175] Calculation results When applied to a system, dynamic control of the system can be achieved.

[0176] S420. Control the error result according to the control quantity to obtain the control parameters for the unmanned vibratory roller.

[0177] like Figure 5 The diagram illustrates the data flow of the unmanned vibratory roller driving control method according to an embodiment of the present invention. Specifically, the positioning device provides the measured values ​​of the vehicle body coordinates, heading angle, and articulation point steering angle. First, the vehicle body position information is processed by an unscented Kalman filter (UKF), and the more accurate vehicle body coordinate estimates are transformed according to the vehicle body dimensions to calculate the current position information of the roller. Second, the desired coordinates obtained after trajectory planning are... , Desired point curvature and expected speed The error is calculated between the estimated value and the actual value; finally, the above data is transmitted to the LQR controller, and the control quantity is... Transmitted to the controlled object.

[0178] In summary, the unmanned vibratory roller driving control method provided by this invention has the following advantages:

[0179] 1) By combining the unscented Kalman filter with the LQR controller, the vehicle state estimation is updated in real time through the unscented Kalman filter, which compensates for the influence of vibration rolling state measurement noise and improves the accuracy of the LQR controller state error input, thereby improving the accuracy and robustness of trajectory tracking.

[0180] 2) The unscented Kalman filter used in this invention does not linearize the system state equation when processing the state estimation of an unmanned vibratory roller. Instead, it uses unscented transformation to infer the posterior probability distribution of the nonlinear process, thus avoiding linearization error and providing a more accurate and reliable state estimate.

[0181] 3) The LQR controller adjusts the dynamic characteristics of the system through the weight matrices Q and R, and can make trade-offs among multiple performance indicators. In practical engineering applications, it can be adjusted according to different requirements for energy consumption and tracking accuracy, and has strong flexibility and adaptability.

[0182] As another embodiment of the present invention, an unmanned vibratory roller driving control device 100 is provided to implement the unmanned vibratory roller driving control method described above, wherein, as Figure 6 As shown, it includes:

[0183] The positioning information acquisition module 110 is used to acquire the positioning information of the unmanned vibratory roller, the positioning information including at least the vehicle body coordinates, heading angle and articulation point turning angle;

[0184] Kalman filter module 120 is used to process the positioning information through an unscented Kalman filter to obtain the real-time state estimate of the unmanned vibratory roller.

[0185] The error calculation module 130 is used to obtain the trajectory planning result of the unmanned vibratory roller, and to perform error calculation based on the trajectory planning result and the real-time state estimate to obtain the error result;

[0186] LQR control module 140 is used to input the error result to the LQR controller to obtain the control parameters of the unmanned vibratory roller;

[0187] The driving control module 150 is used to control the driving of the unmanned vibratory roller according to the control parameters of the unmanned vibratory roller.

[0188] The unmanned vibratory roller driving control device provided by this invention processes the acquired positioning information using an unscented Kalman filter, calculates the error between the planning result and the real-time state estimate output by the unscented Kalman filter, and inputs the error result to an LQR controller to obtain the unmanned vibratory roller control parameters. Based on these control parameters, the unmanned vibratory roller's movement is controlled. This unmanned vibratory roller driving control method combines an unscented Kalman filter with an LQR controller, enabling the unmanned vibratory roller to obtain relatively accurate control error values ​​even when the system measurement parameters contain disturbances, and achieving accurate tracking of the desired trajectory according to the proposed control priority. Therefore, the unmanned vibratory roller driving control device of this invention can improve the robustness of unmanned vibratory roller control and the accuracy of trajectory tracking in complex environments.

[0189] The specific working principle of the unmanned vibratory roller driving control device in this embodiment of the invention can be referred to the description of the unmanned vibratory roller driving control method above, and will not be repeated here.

[0190] As another embodiment of the present invention, an unmanned vibratory roller driving control system 10 is provided, wherein, as Figure 7 As shown, it includes: a positioning device 200, a decision-making device 300, an execution device 400, and the unmanned vibratory roller travel control device 100 mentioned above. The positioning device 200, the decision-making device 300, and the execution device 400 are all communicatively connected to the unmanned vibratory roller travel control device 100.

[0191] The positioning device 200 is used to collect positioning information of the unmanned vibratory roller;

[0192] The decision-making device 300 is used to plan the trajectory of the unmanned vibratory roller and obtain the trajectory planning result.

[0193] The unmanned vibratory roller driving control device 100 is used to process the positioning information according to the unscented Kalman filter, calculate the error according to the trajectory planning result and the real-time state estimate, and input the error result to the LQR controller to obtain the unmanned vibratory roller control parameters.

[0194] The actuator 400 is used to control the movement of the unmanned vibratory roller according to the control parameters of the unmanned vibratory roller.

[0195] In this embodiment of the invention, the positioning device 200 is responsible for determining the precise position and heading of the vibratory roller, including a GPS global positioning system and an inertial measurement unit (IMU); wherein the GPS receiver and IMU are located inside the cab, and the positioning antenna is located at the top of the cab.

[0196] The actuator 400 is responsible for executing the instructions from the operation calculation unit to control the movement and steering of the vibratory roller, including the drive and steering systems. Specifically, the actuator mainly includes a hydraulic steering system and a drive system at the rigid hinge of the vibratory roller. The hydraulic solenoid valves in the steering system receive steering control signals from the operation calculation unit, control the extension and retraction of the left and right hydraulic cylinders, and, combined with feedback information from angle sensors, achieve closed-loop control of the hydraulic cylinder extension and retraction to precisely control the steering angle of the vibratory roller. The drive system includes an engine control module and a brake control module. After receiving the operation instructions from the operation calculation unit, the engine control module uses matching electrical devices to achieve automatic ignition and automatic shutdown, and continuously adjusts the engine throttle based on CAN bus information. The brake control module also receives instructions from the operation calculation unit via the CAN bus, controls the parking solenoid valve through electrical devices, and sends idle or stop commands to the engine control module to achieve stopping, parking, and emergency stop control of the vibratory roller.

[0197] In this embodiment of the invention, the decision-making device 300 is used to realize unmanned vibratory roller trajectory planning, and it may further include an information sensing unit, and perform trajectory planning based on the sensing information of the information sensing unit. Figure 8 As shown, the information perception unit mainly consists of sensors that collect environmental and self-information, including lidar, cameras, and millimeter-wave radar, for identifying and tracking static and dynamic obstacles and assessing road conditions. Angle sensors are used to sense the torsional angle at the hinge between the vehicle body and the roller rigid body. The cameras and lidar in the information perception unit are located at the top of the cab, while the millimeter-wave radar is located at the left and right ends of the front of the roller steel frame. The cameras provide real-time images for background monitoring; the lidar located on the top of the vehicle body provides a longer detection range and a wider field of view; the millimeter-wave radar located at the front of the steel frame detects whether there are vehicles or obstacles ahead, providing support for forward collision warning and close-range protection. This multi-layered perception system ensures that the vibratory roller can accurately perceive its surroundings in different driving scenarios.

[0198] In this embodiment of the invention, the unmanned vibratory roller driving control device 100 acquires the hinge point angle information collected by the angle sensor via the CAN bus; acquires the vehicle body coordinate position of the unmanned vibratory roller measured by GPS via the serial interface; and acquires the vehicle body acceleration and angular velocity information measured by the inertial measurement unit (IMU) via the CAN bus. Specifically, upon receiving an operation command, the vibratory roller enters a standby mode, the drive system remains in a parking state, and the steering system is zeroed to ensure that the angle sensor value is zero, i.e., the vibratory roller hinge point remains straight. Subsequently, during the operation of the unmanned vibratory roller, the drive system and steering system of the execution unit are adjusted in real time based on the received sensing information to ensure the trajectory tracking accuracy of the unmanned vibratory roller roller and the speed stability of the vehicle body.

[0199] In this embodiment of the invention, the unmanned vibratory roller driving control system also includes a power supply device, which provides the necessary electrical energy to various devices within the system. The power supply device is installed inside the cab and is equipped with a power monitoring device. While controlling and protecting the power supply to all electrical equipment, it transmits power consumption data and battery status to the unmanned vibratory roller driving control device 100 via a CAN bus.

[0200] In summary, the unmanned vibratory roller driving control system provided by this invention integrates a decision-making device, a positioning device, an execution device, an unmanned vibratory roller driving control device, and a power supply device, providing a complete supporting platform for unmanned vibratory roller driving control and possessing significant practicality. Through the control method and system of this invention, the unmanned vibratory roller can achieve precise path tracking, obstacle avoidance, speed control, and compaction quality monitoring, significantly improving construction efficiency and quality while reducing labor costs and safety risks.

[0201] The specific working principle of the unmanned vibratory roller driving control system provided by this invention can be referred to the description of the unmanned vibratory roller driving control device above, and will not be repeated here.

[0202] It is understood that the above embodiments are merely exemplary embodiments used to illustrate the principles of the present invention, and the present invention is not limited thereto. For those skilled in the art, various modifications and improvements can be made without departing from the spirit and essence of the present invention, and these modifications and improvements are also considered to be within the scope of protection of the present invention.

Claims

1. A method for controlling the movement of an unmanned vibratory roller, characterized in that, include: The positioning information of the unmanned vibratory roller is obtained, and the positioning information includes at least the vehicle body coordinates, heading angle and articulation point steering angle; The positioning information is processed by an unscented Kalman filter to obtain a real-time state estimate of the unmanned vibratory roller. Obtain the trajectory planning results of the unmanned vibratory roller, and calculate the error between the trajectory planning results and the real-time state estimate to obtain the error result; The error results are input into the LQR controller to obtain the control parameters for the unmanned vibratory roller. The movement of the unmanned vibratory roller is controlled according to the aforementioned unmanned vibratory roller control parameters; The positioning information is processed using an unscented Kalman filter, including: Construct an unscented Kalman filter; The positioning information is input into the unscented Kalman filter for processing to obtain the vehicle body coordinate estimate. Based on the vehicle body size information, the estimated vehicle body coordinates are transformed into coordinate points to obtain the real-time estimated position of the rollers. Constructing an unscented Kalman filter includes: The motion of the unmanned vibratory roller is simplified to planar motion, and the state-space equation of the vibratory roller-vehicle body is established based on the kinematic model. The state transition function is determined based on the vibration roller-vehicle state space equation. The sigma points are predicted based on the state transition function to obtain the Kalman gain and covariance matrix, thus completing the construction of the unscented Kalman filter. The expression for the state-space equation of the vibratory roller-vehicle body is: ， in, Indicates the position and speed of the vehicle's center of gravity. This represents the velocity component of the vehicle body in the X direction in the global coordinate system. This represents the velocity component of the vehicle body in the Y direction in the global coordinate system. This represents the angular velocity of the vehicle's center of mass in the global coordinate system. This indicates the angle difference between the car body and the roller. This represents the distance from the center of mass of the roller to the hinge point. This indicates the distance from the vehicle's center of gravity to the hinge point; The expression for the state transition function is: ， in, State variables , Indicates time, Indicates the sampling time of the discrete system; This represents the X-axis coordinate of the vehicle's center of mass in the global coordinate system. This represents the Y-axis coordinate of the vehicle's center of mass in the global coordinate system. The heading angle represents the center of gravity of the vehicle body.

2. The unmanned vibratory roller driving control method according to claim 1, characterized in that, Based on the state transition function, the sigma point is predicted to obtain the Kalman gain and covariance matrix, including: Initialize to generate sigma points and determine the state estimation vector and covariance matrix at the initialization time; The sigma point is nonlinearly transformed through the state transition function to obtain the predicted state estimate of the sigma point; Calculate the mean and covariance of the predicted state based on the state estimates of the predicted sigma points; The Kalman gain is calculated based on the mean and covariance of the predicted state, and the state estimate and covariance matrix are updated.

3. The unmanned vibratory roller driving control method according to claim 1 or 2, characterized in that, The error results are input to the LQR controller to obtain the control parameters for the unmanned vibratory roller, including: Construct an LQR controller to determine the control variables; The error result is controlled according to the control quantity to obtain the control parameters for the unmanned vibratory roller.

4. The unmanned vibratory roller driving control method according to claim 3, characterized in that, Construct an LQR controller to determine the control variables, including: Construct error-based state-space equations; Determine the cost function and the output weight matrix; Design an LQR controller based on the error-based state-space equation, cost function, and output weight matrix; The control quantity is determined based on the LQR controller.

5. The unmanned vibratory roller driving control method according to claim 4, characterized in that, Determine the cost function and output weight matrix, including: The error-based state-space equation is discretized to obtain a linear time-varying prediction model for the vibrating mill-drum. The LQR cost calculation equation is determined based on the linear time-varying prediction model of the vibratory roller. The output weight matrix is ​​determined based on the LQR cost calculation equation.

6. An unmanned vibratory roller driving control device, used to implement the unmanned vibratory roller driving control method according to any one of claims 1 to 5, characterized in that, include: The positioning information acquisition module is used to acquire the positioning information of the unmanned vibratory roller, and the positioning information includes at least the vehicle body coordinates, heading angle and articulation point steering angle; The Kalman filter module is used to process the positioning information through an unscented Kalman filter to obtain the real-time state estimate of the unmanned vibratory roller. The error calculation module is used to obtain the trajectory planning results of the unmanned vibratory roller, and to perform error calculation based on the trajectory planning results and the real-time state estimate to obtain the error result; The LQR control module is used to input the error results to the LQR controller to obtain the control parameters of the unmanned vibratory roller. The driving control module is used to control the driving of the unmanned vibratory roller according to the control parameters of the unmanned vibratory roller.

7. A driverless vibratory roller driving control system, characterized in that, include: The positioning device, decision-making device, execution device, and unmanned vibratory roller travel control device as described in claim 6, wherein the positioning device, decision-making device, and execution device are all communicatively connected to the unmanned vibratory roller travel control device; The positioning device is used to collect positioning information of the unmanned vibratory roller; The decision-making device is used to plan the trajectory of the unmanned vibratory roller and obtain the trajectory planning result. The unmanned vibratory roller driving control device is used to process the positioning information according to the unscented Kalman filter, calculate the error according to the trajectory planning result and the real-time state estimate, and input the error result to the LQR controller to obtain the unmanned vibratory roller control parameters. The actuator is used to control the movement of the unmanned vibratory roller according to the control parameters of the unmanned vibratory roller.

Images