Closed-loop drift control method and system for autonomous vehicle in uncertain environment

Abstract

The invention relates to the technical field of closed-loop drift control, and particularly discloses a closed-loop drift control method and system for an autonomous vehicle in an uncertain environment, and the method comprises the steps: building a vehicle kinematic model based on a Frenet coordinate system, and constructing an MPC path tracking controller with a preview mechanism; establishing a three-degree-of-freedom vehicle dynamics model with additional yawing moment; the method comprises the following steps: establishing a UniTile-Ctrl tire dynamic model; designing a robust drift controller by combining LQR and integral sliding mode control on the basis of the whole vehicle dynamic model and the tire dynamic model, designing a torque distribution module by taking minimization of distribution error and tire load utilization rate as targets on the basis of ideal rear wheel rotating speed and additional yawing moment output by the robust controller, and outputting a torque distribution model; according to the drifting vehicle tracking control method, closed-loop drifting control can be achieved, the drifting vehicle tracking control performance and robustness are improved, and safety guarantee is provided for limit control over the uncertain road surface of the vehicle.

Description

Technical Field

[0001] This invention relates to the field of closed-loop drift control technology, specifically to a closed-loop drift control method and system for autonomous vehicles in uncertain environments. Background Technology

[0002] With the continuous development of automotive intelligence, more and more vehicle control systems are pursuing higher levels of maneuverability and stability. Existing control technologies aim to keep the vehicle stable within a linear range, ensuring a safer driving experience. However, in more extreme conditions, professional drivers can utilize vehicle control to achieve more extreme maneuvers, such as drifting to achieve high-speed cornering at extreme speeds.

[0003] In current drifting conditions, only professional drivers can perform this extreme maneuver using the handbrake and accelerator. Currently, there are no simple, mass-producible algorithms available for autonomous vehicles specifically designed for this situation. Furthermore, drift controllers require extremely high precision in their state parameters. In uncertain environments, current drift controllers lack robustness in handling uncertainty. Therefore, designing a stable drift controller suitable for real-world driving conditions is essential. Summary of the Invention

[0004] The purpose of this invention is to provide a closed-loop drift control method and system for autonomous vehicles in uncertain environments, so as to solve the problems mentioned in the background art.

[0005] To achieve the above objectives, the present invention provides the following technical solution: A closed-loop drift control method for autonomous vehicles in uncertain environments, the method comprising: A vehicle kinematics model is established based on the Frenet coordinate system, and an MPC path tracking controller with a pre-aiming mechanism is constructed based on the vehicle kinematics model. Establish a three-degree-of-freedom vehicle dynamics model with additional yaw moment; A UniTire-Ctrl tire dynamics model was established, and tire parameters were fitted based on experimental data; Based on the vehicle dynamics model and tire dynamics model, a robust drift controller is designed by combining LQR and integral sliding mode control. Its control variables include front wheel steering angle, rear wheel speed and vehicle additional yaw moment. Based on the ideal rear wheel speed and additional yaw moment output by the robust controller, a torque distribution module is designed to minimize distribution error and tire load utilization, calculate and output actuator-level left and right rear wheel torque commands.

[0006] As a further embodiment of the present invention, the pre-aiming mechanism is implemented by defining a pre-aiming distance d, and the path parameters of the pre-aiming point L1 are calculated by the following formula: ; ; Among them, e L1 s represents the lateral distance at the aiming point. L1 It represents the distance already traveled at the vehicle's reference path aiming point. Let be the heading error at the vehicle's aiming point, e be the lateral distance between the current vehicle and the reference path, and s be the distance the vehicle has traveled along the reference path. This represents the current heading error of the vehicle.

[0007] As a further aspect of the present invention, the extended kinematic model, which includes the current state and the aiming point state, is linearized to obtain a state-space equation for the design of the MPC controller, whose state vector ζ and control vector U... M Defined as: ; ; Among them, V xeq V yeq r eq These represent the drift lateral speed, longitudinal speed, and yaw rate at the drift equilibrium point, respectively.

[0008] As a further embodiment of the present invention, the differential equations of the three-degree-of-freedom vehicle dynamics model are as follows: ; ; ; Where m is the vehicle's own weight, V x Let β be the vehicle's longitudinal speed, β be the vehicle's sideslip angle, r be the vehicle's yaw rate, and δ be the vehicle's front wheel steering angle. For the additional yaw moment of the vehicle, I Z For the vehicle's rotational inertia, F xr and F yr These are the longitudinal and lateral forces on the rear wheels of the vehicle. f F is the distance from the center of gravity to the front wheel. yf It is the lateral force on the front wheels of the vehicle, l r This is the distance from the center of gravity to the rear wheel.

[0009] As a further aspect of the present invention, the robust controller is designed for system models that include parameter uncertainties and external disturbances: ; in, and The unknown time-varying function matrix represents the parameter uncertainty, D is the external disturbance corresponding to the random road surface, A is the state matrix, B is the control input matrix, X is the state space error parameter, and U is the control input.

[0010] As a further embodiment of the present invention, the integral sliding mode control surface is designed as follows: ; ; Where S(t) is the drift integral sliding surface, H is a constant matrix, and simultaneously satisfies K is the feedback matrix of LQR, X(t) is the state-space parameter, and U L (t) represents the controller input of the LQR; The integral sliding mode control law is: ; The final control law of the robust controller is composed of LQR control input and integral sliding mode control input.

[0011] As a further embodiment of the present invention, the torque distribution module distributes the ideal additional yaw moment of the entire vehicle to the left and right rear wheels based on the load transfer caused by vehicle roll: ; ; in, The mass that causes the roll deviation. To add a yaw moment to the vehicle as desired, F zr This refers to the load on the rear axle of the vehicle. , The additional yaw moment is distributed to the left and right rear wheels, respectively.

[0012] As a further embodiment of the present invention, the final torque command for the left and right rear wheels is calculated using the following formula: ; ; in, This is the proportionality coefficient. Let be the moment of inertia of the wheel relative to the axis of rotation. This refers to the wheel speed of the vehicle's left rear wheel. This refers to the wheel speed of the vehicle's right rear wheel. Ideal driving force for the left rear wheel of the vehicle. Ideal driving force for the right rear wheel of the vehicle. R is the ideal rear wheel speed output by the overall drift controller.e The effective radius of the wheel, , These are the ideal rear wheel speeds for the left and right wheels, respectively.

[0013] This invention also provides a closed-loop drift control system for autonomous vehicles in uncertain environments, used to implement the aforementioned closed-loop drift control method for autonomous vehicles in uncertain environments, characterized in that the system comprises: The aiming mechanism module is used to build a vehicle kinematics model based on the Frenet coordinate system, and to construct an MPC path tracking controller with an aiming mechanism based on the vehicle kinematics model. The vehicle dynamics model building module is used to establish a three-degree-of-freedom vehicle dynamics model with additional yaw moment; The tire dynamics model building module is used to build UniTire-Ctrl tire dynamics models and fit tire parameters based on experimental data. The controller design module is used to design a robust drift controller based on the vehicle dynamics model and tire dynamics model, combined with LQR and integral sliding mode control. Its control quantities include front wheel steering angle, rear wheel speed and additional yaw moment of the vehicle. The output module is used to design a torque distribution module based on the ideal rear wheel speed and additional yaw moment output by the robust controller, with the goal of minimizing distribution error and tire load utilization, to calculate and output actuator-level left and right rear wheel torque commands.

[0014] Compared with the prior art, the beneficial effects of the present invention are as follows: The present invention adopts a hierarchical control system structure. The upper layer uses a pre-aiming MPC controller to follow the vehicle's trajectory, obtain the drift equilibrium point based on the path information, and provide the corresponding parameters to the lower layer. The lower layer considers the parameters during tire drift and the road surface uncertainty, and uses the LQR algorithm superimposed with integral sliding mode control as the vehicle controller. The hierarchical control structure separates path tracking and drift state control, reduces the complexity of control, and forms a closed-loop drift control. Attached Figure Description

[0015] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention.

[0016] Figure 1 A flowchart of a closed-loop drift control method for autonomous vehicles under uncertain environments provided in an embodiment of the present invention.

[0017] Figure 2 This is a schematic diagram of a vehicle model for controller design provided in an embodiment of the present invention.

[0018] Figure 3 This is a schematic diagram of the algorithm architecture provided in an embodiment of the present invention. Detailed Implementation

[0019] To make the technical problems to be solved, the technical solutions, and the beneficial effects of the present invention clearer, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the present invention and are not intended to limit the present invention.

[0020] Figure 1 This is a flowchart of a closed-loop drift control method for autonomous vehicles under uncertain environments. In this embodiment of the invention, the closed-loop drift control method for autonomous vehicles under uncertain environments includes: A vehicle kinematic model is established based on the Frenet coordinate system, and a model predictive control (MPC) path tracking controller with a preview mechanism is constructed based on the vehicle kinematic model. Establish a three-degree-of-freedom vehicle dynamics model with additional yaw moment; A UniTire-Ctrl tire dynamics model was established, and tire parameters were fitted based on experimental data; Based on the vehicle dynamics model and tire dynamics model, a robust drift controller is designed by combining a linear quadratic regulator (LQR) and integral sliding mode control. The control quantities include the front wheel steering angle, the rear wheel speed and the additional yaw moment of the whole vehicle. Based on the ideal rear wheel speed and additional yaw moment output by the robust controller, a torque distribution module is designed to minimize distribution error and tire load utilization, calculate and output actuator-level left and right rear wheel torque commands.

[0021] In this embodiment, step one is to establish a pre-aiming MPC path tracking controller, and the path tracking model used is the vehicle kinematics model in the Frenet coordinate system.

[0022] Step 2: Establish a vehicle dynamics model suitable for controller design. Due to the limitations of controller computing power and the accuracy requirements of the vehicle model, this invention adopts a model with additional yaw moment. A three-degree-of-freedom vehicle dynamics model.

[0023] Step 3: Establish a tire dynamics model. The tire dynamics model adopts the UniTire-Ctrl tire model, and the corresponding tire parameters are fitted according to the MatLab toolbox.

[0024] Step 4: Based on the above vehicle dynamics model and tire dynamics model, considering parameter uncertainties and road surface uncertainties, establish a controller that integrates LQR and integral sliding mode, with the control variables being the front wheel steering angle, rear wheel speed, and additional yaw moment of the vehicle.

[0025] Step 5: Take into account the obtained ideal rear wheel speed and additional yaw moment in a unified manner, and design a torque distribution execution module to minimize distribution error, minimize tire load utilization, and take into account actuator constraints, and apply it to the vehicle.

[0026] In a preferred embodiment of the present invention, the steps of establishing a vehicle kinematic model based on the Frenet coordinate system and constructing an MPC path tracking controller with a pre-aiming mechanism based on the vehicle kinematic model are as follows: The vehicle kinematics model is established using the Frenet coordinate system as follows: ; ; ; Where e is the lateral distance between the vehicle and the reference path, and s is the distance the vehicle has traveled along the reference path. V represents the vehicle's heading error. x V y Let r be the vehicle's longitudinal and lateral velocities, β be the vehicle's yaw rate, β be the vehicle's sideslip angle, and k be the vehicle's curvature.

[0027] By incorporating a pre-aiming mechanism, which is implemented by defining a pre-aiming distance d, L1 = s + d can be approximated as the reference path distance at the pre-aiming point. The vehicle kinematics model at the pre-aiming point is as follows: ; ; ; Among them, e L1 s represents the lateral distance at the aiming point. L1 It represents the distance already traveled at the vehicle's reference path aiming point. Let be the heading error at the vehicle's aiming point, e be the lateral distance between the current vehicle and the reference path, and s be the distance the vehicle has traveled along the reference path. V represents the current heading error of the vehicle. x and V y Let k be the longitudinal and lateral velocities of the vehicle. L1 Let be the curvature at the aiming point.

[0028] Differentiating the above equation yields... ; ; Where r is the yaw rate of the vehicle. Linearizing the above kinematic model, its state-space equation is: ; ; ζ represents the vehicle's state-space error, U represents the difference between the actual control input and the path reference input, η represents the control system output, and A represents the path reference input. M and B M These are the system matrix and the input matrix, C. M It is an identity matrix.

[0029] Wherein, system matrix A M and input matrix B M Obtained in the following ways: ; ; ; In a preferred embodiment of the present invention, the extended kinematic model, which includes the current state and the aiming point state, is linearized to obtain the state-space equations used for the design of the MPC controller, whose state vector ζ and control vector U... M Defined as: ; ; Among them, V xeq V yeq r eq These represent the drift lateral speed, longitudinal speed, and yaw rate at the drift equilibrium point, respectively.

[0030] After proposing the system model, it is necessary to construct the MPC optimization problem and solve it in the form of solving the QP problem. The cost function of the controller is shown below: ; in N p and N c It refers to the prediction step size and the control step size, Q. M and R M This is the weight matrix for MPC. At each time step, the goal is to optimize the cost function J to minimize the cost function, at which point we have: ; ; ; At this point, the ideal longitudinal velocity V xd Lateral velocity Vyd yaw rate r d for: .

[0031] like Figure 2 As shown, in a preferred embodiment of the present invention, the differential equations of the three-degree-of-freedom vehicle dynamics model are as follows: ; ; ; Where m is the vehicle's own weight, V x Let β be the vehicle's longitudinal speed, β be the vehicle's sideslip angle, r be the vehicle's yaw rate, and δ be the vehicle's front wheel steering angle. For the additional yaw moment of the vehicle, I Z For the vehicle's rotational inertia, F xr and F yr These are the longitudinal and lateral forces on the rear wheels of the vehicle. f F is the distance from the center of gravity to the front wheel. yf It is the lateral force on the front wheels of the vehicle, l r This is the distance from the center of gravity to the rear wheel.

[0032] In a preferred embodiment of the present invention, the step of establishing the UniTire-Ctrl tire dynamics model and fitting tire parameters based on experimental data includes: Longitudinal slip ratio S x and lateral slip ratio S y for: ; ; Where α is the tire slip angle. R is the wheel speed. e Let V be the effective radius of the wheel and V be the vehicle speed. The normalized longitudinal slip ratio is obtained from this. Lateral slip ratio and combined slip ratio It is given by the following formula: ; ; ; Among them, K x and K y These are the longitudinal slip and lateral stiffness of the tire, respectively, where λ is the direction factor and μ is the longitudinal slip and lateral slip stiffness. x and μ y The adhesion coefficients in the longitudinal and transverse directions, F respectively. z This refers to the vertical force exerted on the tire.

[0033] Normalized dimensionless total tangent and longitudinal force F x and lateral force F y They are respectively: ; ; ; Where E is the curvature factor of the total shear force curve, μ x and μ y Φ represents the longitudinal and transverse adhesion coefficients, respectively, and Φ is the relative overall slip ratio.

[0034] Substituting the vehicle dynamics model into the forward Euler state equations yields matrices A and B. LQR control outputs steering wheel angle commands, wheel speed commands, and additional yaw moment commands, which are then superimposed with integral sliding mode control to enhance the robustness of the control system.

[0035] Considering parameter uncertainties and road surface disturbances, the linearized state equation is established as follows: ; in, and The unknown time-varying function matrix represents the parameter uncertainty, D is the external disturbance corresponding to the random road surface, A is the state matrix, B is the control input matrix, X is the state space error parameter, and U is the control input.

[0036] For an LQR controller, assume the state-space equations are the nominal space equations: ; ; ; Where β d These represent the ideal centroid sideslip angle of the vehicle under the path tracking controller. , For the LQR controller, the desired rear wheel speed, , and These are the front wheel steering angle, rear wheel speed, and additional yaw moment of the vehicle under drift equilibrium conditions.

[0037] Redesign the LQR drift controller, where the quadratic optimization function of the LQR drift controller is: ; Among them, Q L and R L These are the weight matrices for the state variables and the control variables, respectively. ; ; in They are respectively the corresponding The weight, They are respectively the corresponding The weight.

[0038] By solving the equations By finding P, the feedback matrix can be obtained. The desired control quantity is: .

[0039] In a preferred embodiment of the present invention, the integral sliding mode control surface is designed as follows: ; ; Where S(t) is the drift integral sliding surface, H is a constant matrix, and simultaneously satisfies K is the feedback matrix of the LQR controller, X(t) is the state-space parameter, and U L (t) represents the controller input of the LQR; Taking the derivative of the above equation, we can obtain: ; make The integral sliding mode control law can be obtained as follows: ; The final control law of the robust controller is composed of a combination of LQR control input and integral sliding mode control input; in This is the control input for the integral sliding mode controller. The control law for the overall robust drift controller can then be obtained as follows: ; ; In the formula, All are normal numbers.

[0040] The steering wheel angle command directly controls the steering wheel angle of the vehicle, while the wheel speed command is combined with the current wheel speed of the vehicle to control the wheel speed. At the same time, an additional yaw moment is superimposed, and finally the rear wheel torque command is output to control the rear wheel torque of the vehicle.

[0041] In a preferred embodiment of the present invention, the step of designing a torque distribution module based on the ideal rear wheel speed and additional yaw moment output by the robust controller, with the goal of minimizing distribution error and tire load utilization, and calculating and outputting actuator-level left and right rear wheel torque commands, is as follows: When calculating the rear wheel torque during a vehicle drift, the vehicle's weight and lateral tilt should be taken into account. The ideal rear wheel speeds for both left and right wheels at this time are determined accordingly. , The differences are as follows: ; ; Where d is the width of the vehicle body. The ideal rear wheel speed output by the overall drift controller, at which point the mass exhibiting roll deviation... : ; ; ; Where h is the height of the vehicle's center of gravity, F zr This refers to the load on the rear axle of the vehicle. Ideal driving force for the left rear wheel of the vehicle. Ideal driving force for the right rear wheel of the vehicle. This is the rear axle driving force during vehicle drift balancing. It also considers the expected additional yaw moment for the entire vehicle. The additional yaw moment distribution is defined by the load on the left and right wheels. The torque distribution module distributes the ideal additional yaw moment of the entire vehicle to the left and right rear wheels based on the load transfer caused by vehicle roll. ; ; in, The mass that causes the roll deviation. To add a yaw moment to the vehicle as desired, F zr This refers to the load on the rear axle of the vehicle. , The additional yaw moment is distributed to the left and right rear wheels, respectively.

[0042] In a preferred embodiment of the present invention, the final torque command for the left and right rear wheels is calculated using the following formula: ; ; in, This is the proportionality coefficient. Let be the moment of inertia of the wheel relative to the axis of rotation. This refers to the wheel speed of the vehicle's left rear wheel. This refers to the wheel speed of the vehicle's right rear wheel. Ideal driving force for the left rear wheel of the vehicle. Ideal driving force for the right rear wheel of the vehicle. R is the ideal rear wheel speed output by the overall drift controller. e The effective radius of the wheel, , These are the ideal rear wheel speeds for the left and right wheels, respectively.

[0043] Based on the front wheel steering angle command and the rear wheel torque command, the vehicle is controlled to complete closed-loop tracking drift. The overall algorithm architecture diagram of this invention is as follows: Figure 3 As shown.

[0044] This invention employs a hierarchical control system structure. The upper layer uses a pre-aiming MPC controller to follow the vehicle's trajectory, obtain the drift equilibrium point based on path information, and provide corresponding parameters to the lower layer. The lower layer considers tire drift parameters and road surface uncertainties, using an LQR algorithm superimposed with integral sliding mode control as the vehicle controller. Based on the parameters provided by the upper-layer path tracking controller and the current vehicle state, it calculates the front wheel steering angle, rear wheel drive torque, and additional yaw moment as vehicle control variables. The lower layer considers vehicle weight, roll offset, minimizing tire compliance utilization, and actuator constraints to obtain the torque of the left and right wheels. This hierarchical control structure separates path tracking and drift state control, reducing control complexity and forming a closed-loop drift control, providing a technical solution for closed-loop drift control.

[0045] The present invention also provides a closed-loop drift control system for autonomous vehicles in uncertain environments, used to implement the aforementioned closed-loop drift control method for autonomous vehicles in uncertain environments, the system comprising: The aiming mechanism module is used to build a vehicle kinematics model based on the Frenet coordinate system, and to construct an MPC path tracking controller with an aiming mechanism based on the vehicle kinematics model. The vehicle dynamics model building module is used to establish a three-degree-of-freedom vehicle dynamics model with additional yaw moment; The tire dynamics model building module is used to build UniTire-Ctrl tire dynamics models and fit tire parameters based on experimental data. The controller design module is used to design a robust drift controller based on the vehicle dynamics model and tire dynamics model, combined with LQR and integral sliding mode control. Its control quantities include front wheel steering angle, rear wheel speed and additional yaw moment of the vehicle. The output module is used to design a torque distribution module based on the ideal rear wheel speed and additional yaw moment output by the robust controller, with the goal of minimizing distribution error and tire load utilization, to calculate and output actuator-level left and right rear wheel torque commands.

[0046] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A closed-loop drift control method for autonomous vehicles under uncertain environments, characterized in that, The method includes: A vehicle kinematics model is established based on the Frenet coordinate system, and an MPC path tracking controller with a pre-aiming mechanism is constructed based on the vehicle kinematics model. Establish a three-degree-of-freedom vehicle dynamics model with additional yaw moment; A UniTire-Ctrl tire dynamics model was established, and tire parameters were fitted based on experimental data; Based on the vehicle dynamics model and tire dynamics model, a robust drift controller is designed by combining LQR and integral sliding mode control. Its control variables include front wheel steering angle, rear wheel speed and vehicle additional yaw moment. Based on the ideal rear wheel speed and additional yaw moment output by the robust controller, a torque distribution module is designed to minimize distribution error and tire load utilization, calculate and output actuator-level left and right rear wheel torque commands.

2. The closed-loop drift control method for autonomous vehicles under uncertain environments according to claim 1, characterized in that, The pre-aiming mechanism is implemented by defining a pre-aiming distance d, and the path parameters of the pre-aiming point L1 are calculated using the following formula: ； ； Among them, e L1 s represents the lateral distance at the aiming point. L1 It represents the distance already traveled at the vehicle's reference path aiming point. Let be the heading error at the vehicle's aiming point, e be the lateral distance between the current vehicle and the reference path, and s be the distance the vehicle has traveled along the reference path. This represents the current heading error of the vehicle.

3. The closed-loop drift control method for autonomous vehicles under uncertain environments according to claim 2, characterized in that, The extended kinematic model, which includes the current state and the aiming point state, is linearized to obtain the state-space equations for the design of the MPC controller, with its state vector ζ and control vector U. M Defined as: ； ； Among them, V xeq V yeq r eq These represent the drift lateral speed, longitudinal speed, and yaw rate at the drift equilibrium point, respectively.

4. The closed-loop drift control method for autonomous vehicles under uncertain environments according to claim 3, characterized in that, The differential equations of the three-degree-of-freedom vehicle dynamics model are as follows: ； ； ； Where m is the vehicle's own weight, V x Let β be the vehicle's longitudinal speed, β be the vehicle's sideslip angle, r be the vehicle's yaw rate, and δ be the vehicle's front wheel steering angle. For the additional yaw moment of the vehicle, I Z For the vehicle's rotational inertia, F xr and F yr These are the longitudinal and lateral forces on the rear wheels of the vehicle. f F is the distance from the center of gravity to the front wheel. yf It is the lateral force on the front wheels of the vehicle, l r This is the distance from the center of gravity to the rear wheel.

5. The closed-loop drift control method for autonomous vehicles under uncertain environments according to claim 4, characterized in that, The robust controller is designed for system models that include parameter uncertainties and external disturbances. ； in, and The unknown time-varying function matrix represents the parameter uncertainty, D is the external disturbance corresponding to the random road surface, A is the state matrix, B is the control input matrix, X is the state space error parameter, and U is the control input.

6. The closed-loop drift control method for autonomous vehicles under uncertain environments according to claim 5, characterized in that, The integral sliding mode control surface is designed as follows: ； ； Where S(t) is the drift integral sliding surface, H is a constant matrix, and simultaneously satisfies K is the feedback matrix of LQR, X(t) is the state-space parameter, and U L (t) represents the controller input of the LQR; The integral sliding mode control law is: ； The final control law of the robust controller is composed of LQR control input and integral sliding mode control input.

7. The closed-loop drift control method for autonomous vehicles under uncertain environments according to claim 6, characterized in that, The torque distribution module distributes the ideal additional yaw moment of the entire vehicle to the left and right rear wheels based on the load transfer caused by vehicle roll: ； ； in, The mass that causes the roll deviation. To add a yaw moment to the vehicle as desired, F zr This refers to the load on the rear axle of the vehicle. , The additional yaw moment is distributed to the left and right rear wheels, respectively.

8. The closed-loop drift control method for autonomous vehicles under uncertain environments according to claim 7, characterized in that, The final torque command for the left and right rear wheels is calculated using the following formula: ； ； in, This is the proportionality coefficient. Let be the moment of inertia of the wheel relative to the axis of rotation. This refers to the wheel speed of the vehicle's left rear wheel. This is the wheel speed of the vehicle's right rear wheel. Ideal driving force for the left rear wheel of the vehicle. Ideal driving force for the right rear wheel of the vehicle. R is the ideal rear wheel speed output by the overall drift controller. e The effective radius of the wheel, , These are the ideal rear wheel speeds for the left and right wheels, respectively.

9. A closed-loop drift control system for autonomous vehicles in uncertain environments, used to implement the closed-loop drift control method for autonomous vehicles in uncertain environments as described in any one of claims 1-8, characterized in that, The system includes: The aiming mechanism module is used to build a vehicle kinematics model based on the Frenet coordinate system, and to construct an MPC path tracking controller with an aiming mechanism based on the vehicle kinematics model. The vehicle dynamics model building module is used to build a three-degree-of-freedom vehicle dynamics model with additional yaw moment; The tire dynamics model building module is used to build UniTire-Ctrl tire dynamics models and fit tire parameters based on experimental data. The controller design module is used to design a robust drift controller based on the vehicle dynamics model and tire dynamics model, combined with LQR and integral sliding mode control. Its control quantities include front wheel steering angle, rear wheel speed and additional yaw moment of the vehicle. The output module is used to design a torque distribution module based on the ideal rear wheel speed and additional yaw moment output by the robust controller, with the goal of minimizing distribution error and tire load utilization, to calculate and output actuator-level left and right rear wheel torque commands.

Images