A 4WID-4WIS vehicle path tracking control method and system based on the HJI theory

Abstract

The present invention discloses a 4WID-4WIS vehicle path tracking control method and system based on the HJI theory. The system includes five modules: information perception, human-machine interaction, steering control, state monitoring, and lower-layer execution. The information perception module is used to obtain vehicle state parameters. The present invention designs a corner control law and an adaptive control law by using the Hamilton-Jacobi-Issacs inequality, and attenuates the influence of the approximation error interference on the system to below the specified level. The steering control module obtains the control laws for the front and rear wheel angles of the vehicle, and at the same time designs a radial basis neural network algorithm to approximate the unknown interference of the system. A suitable Lyapunov function and an interference suppression evaluation index are designed. The state monitoring module monitors the path condition and vehicle stability in real time, and when the vehicle state is abnormal, it transmits the information to the human-machine interaction module in a timely manner, and at the same time starts the direct yaw moment control to ensure the safety of vehicle driving. The present invention can effectively reduce the influence of unknown interference on the system.

Description

Technical Field

[0001] The present invention relates to the technical field of vehicle dynamics control, and mainly relates to a path tracking control method and system for a 4WID-4WIS vehicle based on the HJI theory. Background Technique

[0002] A four-wheel independent steering and independent drive (4WIS-4WID) vehicle based on in-wheel motors and steering motors is an advanced vehicle with multiple actuators and multiple control degrees of freedom. It can achieve four-wheel independent drive, braking, and steering, and has unique advantages in realizing the coordinated and integrated control of various vehicle active safety technologies. It is one of the important directions for future intelligent electric vehicles.

[0003] At present, four-wheel steering technology is mostly used for vehicle active safety control to achieve vehicle lateral stability. For example, in Chinese Patent ZL201911423195X, four-wheel steering technology is used to control the vehicle's sideslip angle and lateral acceleration by using a data-driven model predictive control method to achieve vehicle driving stability. However, there is still little research on the path tracking control of 4WIS-4WID vehicles, and their advantages in realizing autonomous driving have not been fully utilized.

[0004] The path tracking control of intelligent vehicles refers to how to control the vehicle to travel along the planned path and ensure the driving safety and ride comfort of the vehicle. Domestic and foreign scholars have conducted a large number of studies on two-wheel steering vehicles and proposed many advanced control methods such as optimal control, robust control, adaptive control, and sliding mode control. However, due to the complex longitudinal and lateral dynamics mechanism of the vehicle and the defects of various control methods themselves, there are still many unsolved problems in the path tracking control of vehicles at present, such as large computational complexity of the control algorithm, time-varying uncertainty and unknown disturbances of the vehicle model, etc. Summary of the Invention

[0005] Object of the Invention: Aiming at the problems existing in the above background technique, the present invention uses a radial basis neural network RBFNN to approximate the unknown disturbance term of the system, and at the same time uses the Hamilton-Jacobi-Issacs (HJI) inequality to design a steering controller to improve the robustness of the system. A path tracking control system and method for a 4WID / 4WIS vehicle based on the HJI theory are proposed. The specific technical solutions are as follows:

[0006] Technical Solution: To achieve the above object, the technical solution adopted by the present invention is:

[0007] A path tracking control method for a 4WID-4WIS vehicle based on the HJI theory, comprising the following steps:

[0008] Step S1, establishing a second-order dynamic equation of the tracking error variable based on the 4WID-4WIS vehicle two-degree-of-freedom lateral dynamics model;

[0009] Step S2: establishing a control system state space model according to the dynamic equation;

[0010] Step S3: establishing a radial basis function neural network algorithm, which performs real-time approximation on unknown interference terms in the system state space model to improve the tracking performance of the control system;

[0011] Step S4: Establish performance indicators for judging the interference suppression capability of the system;

[0012] Step S5: Design the Lyapunov function and the disturbance rejection evaluation index, and design the front and rear wheel steering angle control law and the radial basis function neural network adaptive control law based on the HJI inequality.

[0013] Furthermore, the kinetic equation in step S1 is expressed as follows:

[0014]

[0015] where e=[e1,e2] T , e1 is the vehicle's current position error, e2 is the vehicle's current direction error; u=[δ f ,δ r ] T , where δ f is the front wheel turning angle, δ r is the rear wheel turning angle; is the theoretical direction change rate, d is the unknown external interference term of the system, A, A1, B, B1 are the system parameter matrices, as follows:

[0016]

[0017]

[0018] Among them, k1 and k2 are the vehicle's front and rear wheel cornering stiffnesses respectively, a and b are the distances from the vehicle's center of mass to the front axle and rear axle respectively, m is the vehicle's mass, and I z is the moment of inertia of the vehicle around the Z axis, u is the longitudinal velocity of the vehicle, and R is the curvature radius of the target path.

[0019] Furthermore, in step S2, the state variables are first defined as follows:

[0020]

[0021] The state space model of the control system is established as follows:

[0022]

[0023] Among them

[0024] Furthermore, the radial basis neural network includes an input layer, a hidden layer, and an output layer; where:

[0025] The input vector of the input layer is x = [x1, x2] T , where x1 and x2 are state variables;

[0026] The hidden layer includes 5 neurons, and the output is h(x) = [h i T , i = 1, 2, 3, 4, 5, where the output of the i-th neuron h i is:

[0027]

[0028] Among them, c i is the coordinate vector of the center point of the Gaussian basis function of the i-th neuron in the hidden layer, b = [b i T , b i is the width of the Gaussian basis function of the i-th neuron in the hidden layer;

[0029] The output layer outputs the interference approximation value W is the weight matrix of the neural network, and there exists an ideal interference approximation value satisfying:

[0030]

[0031] Among them, W * is the ideal weight matrix of the neural network, and ε0 is a given minimum value;

[0032] Define the approximation error The weight error The model uncertain interference is expressed as follows:

[0033] d = W * h(x) + ε d

[0034] Furthermore, the interference suppression performance index in step S4 is expressed as follows:

[0035]

[0036] Among them, z = px1, p is a given constant, and the smaller J is, the better the system robustness is.

[0037] Furthermore, the front and rear wheel steering angle control laws in step S5 are expressed as follows:

[0038] ​​

[0039] The radial basis neural network adaptive control law is expressed as follows:

[0040]

[0041] where \(k_2\), \(\eta\), and \(\mu\) are all given positive constants;

[0042] Define the Lyapunov function as follows:

[0043]

[0044] Then:

[0045] Let

[0046] Also

[0047] It can be obtained that:

[0048]

[0049] Substitute the front and rear wheel steering angle control law and the radial basis neural network adaptive control law:

[0050]

[0051] It can be proved that the control law and the adaptive law satisfy:

[0052]

[0053] According to the HJI principle, \(J\leq\eta\), and the designed system is stable.

[0054] A vehicle path tracking control system using the above 4WID-4WIS vehicle path tracking control method based on the HJI theory includes an information perception module, a steering control module, a state monitoring module, a lower-level execution module, and a human-computer interaction module;

[0055] The information perception module collects state information including road information and vehicle information, and transmits it to the steering control module and the state monitoring module;

[0056] The steering control module obtains the front and rear wheel steering angle control law and the radial basis neural network adaptive control law of the vehicle based on the HJI theory according to the state information and the target path, and transmits the control law and the adaptive law to the lower-level execution module to realize the automatic path tracking of the vehicle;

[0057] The state monitoring sub-module includes a path tracking monitoring sub-module and a vehicle stability monitoring sub-module; the path tracking monitoring sub-module obtains vehicle state information and determines whether the current position error and direction error of the vehicle exceed a preset safety threshold; when the vehicle position error or direction error exceeds the preset safety threshold, the driver is prompted to take over the vehicle through the human-machine interaction module, and at the same time, the automatic path tracking system is terminated; where the position error refers to the distance from the vehicle's center of mass to the center line of the target path, and the direction error refers to the vehicle direction error relative to the target path.

[0058] The vehicle stability monitoring sub-module receives vehicle state information, estimates the vehicle's center of mass sideslip angle in real time, and determines whether the current estimated value of the vehicle's center of mass sideslip angle exceeds a preset critical value of the center of mass sideslip angle; when the center of mass sideslip angle exceeds the critical value, the driver is prompted to take over the vehicle and the automatic path tracking system is terminated through the human-machine interaction module, and at the same time, the direct yaw moment control is started.

[0059] The lower-level execution module receives the front and rear wheel steering angle control laws transmitted by the steering control module, controls the front and rear wheel steering motors to achieve path tracking; at the same time, it receives the target longitudinal torques of the four wheels transmitted by the state monitoring module, and controls the hub motors to output the target torques to achieve vehicle stability control.

[0060] Further, the information perception module includes an on-vehicle lidar, a yaw rate sensor, a lateral acceleration sensor, and a longitudinal vehicle speed sensor; the collected state information includes: vehicle lateral position, vehicle direction angle, yaw rate, lateral acceleration, and road environment.

[0061] The vehicle stability monitoring sub-module receives the vehicle yaw rate and lateral acceleration transmitted by the information perception module, and estimates the vehicle's center of mass sideslip angle in real time.

[0062] Further, the stability monitoring sub-module receives the vehicle yaw rate and lateral acceleration a y , and estimates the vehicle's center of mass sideslip angle β in real time. According to whether the vehicle's center of mass sideslip angle β exceeds the preset critical value β m of the center of mass sideslip angle, it is determined whether the vehicle is critically unstable. If it is critically unstable, the driver is prompted to take over the vehicle and the automatic path tracking system is terminated through the human-machine interaction module, and at the same time, the direct yaw moment control is started to generate an additional yaw moment ΔM according to the designed yaw moment controller, and then it is distributed and converted into the target longitudinal torques T1, T2, T3, and T4 of the wheels, and transmitted to the lower-level execution module.

[0063] The vehicle's center of mass sideslip angle is estimated using the Kalman filter technique, and the algorithm is as follows:

[0064]

[0065] Among them, the state value Observation value A is the state matrix, Q is the process noise covariance matrix, R is the observation noise covariance matrix, P is the estimation error covariance matrix and needs to be iteratively updated, H is the observation matrix and

[0066] The yaw moment controller adopts fuzzy PID control and outputs an additional yaw moment:

[0067]

[0068] Among them, e β =β-β d , β d =0, K P is the proportional gain, K I is the integral gain, K D is the derivative gain, which is a predetermined value (K′ P , K′ I , K′ D ) and the fuzzy correction value (ΔK P , ΔK I , ΔK D ) superposition, that is:

[0069]

[0070] The torque distribution adopts average distribution, and the longitudinal target torque can be obtained by the following formula:

[0071]

[0072] Among them, T d is the driving torque, r is the wheel radius, and d is the vehicle width.

[0073] Beneficial effects:

[0074] The 4WID-4WIS vehicle path tracking control method and system based on the HJI theory provided by the present invention make full use of the unique advantages of four-wheel independent drive-independent steering vehicles, effectively realize vehicle path tracking control, and at the same time coordinate direct yaw moment control, taking into account vehicle driving stability. At the same time, aiming at the time-varying uncertainty interference existing in the system model, the present invention uses a radial basis neural network to approximate it, improving the tracking performance of the control system; using the HJI principle to design the front and rear wheel corner control laws and adaptive laws, which has strong interference suppression ability, can attenuate the influence of the approximation error interference on the system to below the specified level, has good robustness, and the control algorithm has a low complexity and a small amount of calculation. Description of the drawings

[0075] Figure 1 It is the lateral dynamics model diagram of 4WID-4WIS vehicle path tracking in the embodiment of the present invention;

[0076] Figure 2 It is the structural schematic diagram of the 4WID-4WIS vehicle path tracking control system provided by the present invention;

[0077] Figure 3 It is the schematic flow diagram of the 4WID-4WIS vehicle path tracking steering control method in the embodiment of the present invention;

[0078] Figure 4 It is the structural schematic diagram of the radial basis neural network in the embodiment of the present invention. Specific embodiments

[0079] The present invention will be further described in detail below with reference to the accompanying drawings. Obviously, the described embodiments are part of the embodiments of the present invention, rather than all of them. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts fall within the scope of protection of the present invention.

[0080] As Figure 1 shown is the simulated lateral dynamics model diagram of 4WID-4WIS vehicle path tracking in this embodiment. The coordinate system XOY is defined as the ground coordinate system with the lane center line as the target path, and xoy is the vehicle coordinate system. e1 is the current position error of the vehicle, that is, the lateral distance between the vehicle's center of mass and the lane center line, is the current direction angle of the vehicle, is the current ideal direction angle of the vehicle.

[0081] The following specifically introduces the 4WID-4WIS vehicle path tracking control system based on the HJI theory adopted by the present invention, and its structure is as Figure 2 shown, including an information perception module, a steering control module, a state monitoring module, a lower-layer execution module, and a human-machine interaction module.

[0082] The information perception module includes an on-vehicle lidar, a yaw rate sensor, a lateral acceleration sensor, and a longitudinal vehicle speed sensor. The collected state information includes: vehicle lateral position, vehicle direction angle, yaw rate, lateral acceleration, and road environment. And it is transmitted to the steering control module and the state monitoring module.

[0083] The steering control module obtains the control laws of the front and rear wheel steering angles and the radial basis neural network adaptive control law of the vehicle based on the HJI theory according to the state information and the target path, and transmits the control laws and the adaptive laws to the lower-layer execution module to realize the automatic path tracking of the vehicle.

[0084] The state monitoring sub-module includes a path tracking monitoring sub-module and a vehicle stability monitoring sub-module.

[0085] The path tracking monitoring sub-module determines whether the current position error e1 and the direction error e2 of the vehicle exceed the preset error safety threshold e 1m 、e 2m . If it exceeds, the driver is prompted to take over the vehicle through the human-machine interaction module and the automatic path tracking function is terminated. The position error refers to the distance from the vehicle's center of mass to the center line of the target path, and the direction error refers to the vehicle's direction error relative to the target path.

[0086] The stability monitoring sub-module receives the vehicle's yaw rate and the lateral acceleration a y transmitted by the information perception module, and estimates the vehicle's sideslip angle β of the center of mass in real time. According to whether the vehicle's sideslip angle β of the center of mass exceeds the preset critical value β of the sideslip angle of the center of mass m , it is judged whether the vehicle is critically unstable. If it is critically unstable, the driver is prompted to take over the vehicle through the human-machine interaction module and the automatic path tracking system is terminated. At the same time, the direct yaw moment control is started to generate an additional yaw moment ΔM according to the designed yaw moment controller, and then it is distributed and converted into the target longitudinal moments T1, T2, T3, T4 of the wheels and transmitted to the lower-layer execution module.

[0087] The vehicle's sideslip angle of the center of mass is estimated using the Kalman filtering technique, and the algorithm is as follows:

[0088]

[0089] Among them, the state value Observed value A is the state matrix, Q is the process noise covariance matrix, R is the observation noise covariance matrix, P is the estimation error covariance matrix and needs to be iteratively updated, H is the observation matrix and

[0090] The yaw moment controller adopts fuzzy PID control and outputs an additional yaw moment:

[0091]

[0092] Among them, e β =β - β d , β d =0, K P is the proportional gain, K I is the integral gain, K D is the differential gain, which is the sum of the predetermined values (K′ P , K′ I , K′ D ) and the fuzzy correction values (ΔK P , ΔK I , ΔK D ), that is:

[0093]

[0094] The moment distribution adopts average distribution, and the longitudinal target moment can be obtained by the following formula:

[0095]

[0096] where, T d is the driving torque, r is the wheel radius, and d is the vehicle width.

[0097] The lower-layer execution module receives the front and rear wheel corner control laws transmitted by the steering control module, controls the front and rear wheel steering motors to achieve path tracking; at the same time, it receives the target longitudinal moments of the four wheels transmitted by the state monitoring module, controls the hub motors to output the target moments, and realizes vehicle stability control.

[0098] The human-machine interaction module interacts with the driver through the in-vehicle display to start and terminate (ON / OFF) the path tracking function.

[0099] As Figure 3 shown, this embodiment also provides a 4WID-4WIS vehicle path tracking control method based on the HJI theory, which is used to provide the front and rear wheel corner control laws and the radial basis neural network adaptive control laws for the steering control module. The specific steps are as follows:

[0100] Step S1: Establish the second-order dynamic equation of the tracking error variable according to the two-degree-of-freedom lateral dynamics model of the 4WID-4WIS vehicle as follows:

[0101]

[0102] where e = [e1, e2] T , e1 is the vehicle's current position error, and e2 is the vehicle's current direction error. u = [δ f , δ r T , where δ f is the front wheel corner, and δ r is the rear wheel corner. is the theoretical direction change rate, d is the unknown external disturbance term of the system, and A, A1, B, B1 are the system parameter matrices, which are specifically as follows:

[0103]

[0104]

[0105] where, k1 and k2 are the front and rear wheel cornering stiffnesses of the vehicle in sequence, a and b are the distances from the vehicle's center of mass to the front axle and rear axle in sequence, m is the vehicle's total mass, I​z is the moment of inertia of the whole vehicle rotating about the Z-axis, u is the longitudinal speed of the vehicle, and R is the curvature radius of the target path.

[0106] Step S2: Establish a control system state space model according to the dynamic equation.

[0107] First, define the state variables as follows:

[0108]

[0109] Establish the control system state space model as follows:

[0110]

[0111] where

[0112] Step S3: Establish a radial basis neural network algorithm. The radial basis neural network algorithm approximates the unknown disturbance term in the system state space model in real time to improve the tracking performance of the control system. The specific structure is as Figure 4 shown.

[0113] The radial basis neural network includes an input layer, a hidden layer, and an output layer. Among them:

[0114] The input vector of the input layer is x = [x1, x2] T , where x1 and x2 are the state variables defined in step S2;

[0115] The hidden layer includes 5 neurons, and the output is h(x) = [h i T , i = 1, 2, 3, 4, 5, where the output of the i-th neuron h i is:

[0116]

[0117] where, c i is the coordinate vector of the center point of the Gaussian function of the i-th neuron in the hidden layer, b = [b i T , b i is the width of the Gaussian function of the i-th neuron in the hidden layer.

[0118] The output layer outputs the disturbance approximation value W is the weight matrix of the neural network, and there is an ideal disturbance approximation value that satisfies:

[0119]

[0120] where W * is the ideal neural network weight matrix, and ε0 is a given minimum value.

[0121] Define the approximation error Weight error The model uncertainty disturbance is expressed as follows:

[0122] d = W * h(x)+ε d

[0123] Step S4. Establish the performance index for evaluating the disturbance rejection ability of the system as follows:

[0124]

[0125] where z = px1, p is a given constant. The smaller J is, the better the robustness of the system is.

[0126] Step S5. Design the Lyapunov function and the disturbance rejection evaluation index, and design the front and rear wheel steering angle control law and the radial basis neural network adaptive control law based on the HJI inequality.

[0127] The front and rear wheel steering angle control law is expressed as follows:

[0128]

[0129] The radial basis neural network adaptive control law is expressed as follows:

[0130]

[0131] where k2, η, μ are all given positive constants.

[0132] Define the Lyapunov function as follows:

[0133]

[0134] Then:

[0135] Let

[0136] Also

[0137] It can be obtained that:

[0138]

[0139] Substitute the front and rear wheel steering angle control law and the radial basis neural network adaptive control law:

[0140]

[0141] It can be proved that the control law and the adaptive law satisfy:

[0142]

[0143] According to the HJI principle, J ≤ η, and the designed system is stable.

[0144] The above are only the preferred embodiments of the present invention. It should be noted that for those of ordinary skill in the art, without departing from the principle of the present invention, several improvements and refinements can be made, and these improvements and refinements should also be regarded as the protection scope of the present invention.

Claims

1. A 4WID-4WIS vehicle path tracking control method based on the HJI theory, characterized in that, It includes the following steps: Step S1: Establish a second-order dynamic equation of the tracking error variable according to the 4WID-4WIS vehicle two-degree-of-freedom lateral dynamic model; Step S2: Establish a control system state space model according to the dynamic equation; Step S3: Establish a radial basis neural network algorithm, which real-time approximates the unknown disturbance term in the system state space model to improve the tracking performance of the control system; Step S4: Establish a performance index for evaluating the disturbance rejection ability of the system; Step S5: Design a Lyapunov function and a disturbance rejection evaluation index, and design the front and rear wheel steering angle control law and the radial basis neural network adaptive control law based on the HJI inequality; The dynamic equation in step S1 is expressed as follows: where \(e = [e_1, e_2]\) T , \(e_1\) is the vehicle's current position error, and \(e_2\) is the vehicle's current direction error; \(u = [\delta\) f , \(\delta\) r T , where \(\delta\) f is the front wheel steering angle, and \(\delta\) r is the rear wheel steering angle; is the theoretical direction change rate, \(d\) is the unknown external disturbance term acting on the system, and \(A\), \(A_1\), \(B\), \(B_1\) are system parameter matrices, specifically as follows:​ where k1 and k2 are the cornering stiffnesses of the front and rear wheels of the vehicle, a and b are the distances from the vehicle's center of mass to the front and rear axles respectively, m is the total vehicle mass, I z is the moment of inertia of the vehicle about the Z-axis, u is the longitudinal speed of the vehicle, and R is the curvature radius of the target path.

2. The 4WID-4WIS vehicle path tracking control method based on the HJI theory according to claim 1, wherein In step S2, the state variables are first defined as follows: The control system state space model is established as follows: Among them 3. The 4WID-4WIS vehicle path tracking control method based on the HJI theory according to claim 2, characterized in that, The radial basis neural network includes an input layer, a hidden layer, and an output layer; among them: The input vector of the input layer is x = [x1, x2] T , where x1 and x2 are state variables; The hidden layer consists of 5 neurons, and the output is h(x) = [h i T , where i = 1, 2, 3, 4, 5, and the output of the i-th neuron h i is:​ Among them, c i is the coordinate vector of the center point of the Gaussian function of the i-th neuron in the hidden layer, b = [b i T , and b i is the width of the Gaussian function of the i-th neuron in the hidden layer;​ The output layer outputs an interference approximation value W is the weight matrix of the neural network, and there is an ideal interference approximation value Satisfy: Among which, W * is the weight matrix of the ideal neural network, and ε0 is the given minimum value; Define the approximation error Weight error The model uncertainty disturbance is expressed as follows: d = W * h(x) + ε d。 4. The 4WID-4WIS vehicle path tracking control method based on the HJI theory according to claim 1, characterized in that, The disturbance rejection performance index in step S4 is expressed as follows: Where z = px1, p is a given constant. When J is smaller, the system robustness is better.

5. The 4WID-4WIS vehicle path tracking control method based on the HJI theory according to claim 4, characterized in that The front and rear wheel steering angle control law in step S5 is expressed as follows: The radial basis neural network adaptive control law is expressed as follows: Where k2, η, μ are all given positive constants; Define the Lyapunov function as follows: Then: Let Also It can be obtained that: Substitute the front and rear wheel steering angle control law and the radial basis neural network adaptive control law: It can be proved that the control law and the adaptive law satisfy: According to the HJI principle, J ≤ η, and the designed system is stable.

6. A vehicle path tracking control system adopting the 4WID-4WIS vehicle path tracking control method based on the HJI theory according to any one of claims 1-5, characterized in that, It includes an information perception module, a steering control module, a state monitoring module, a lower-layer execution module, and a human-machine interaction module; The information perception module collects state information including road information and vehicle information, and transmits it to the steering control module and the state monitoring module; The steering control module obtains the front and rear wheel steering angle control law and the radial basis neural network adaptive control law of the vehicle based on the HJI theory according to the state information and the target path, and transmits the control law and the adaptive law to the lower-layer execution module to realize the automatic path tracking of the vehicle; The state monitoring sub-module includes a path tracking monitoring sub-module and a vehicle stability monitoring sub-module; the path tracking monitoring sub-module obtains the vehicle state information and judges whether the current position error and direction error of the vehicle exceed the preset safety threshold; When the vehicle position error or direction error exceeds the preset safety threshold, the driver is prompted to take over the vehicle through the human-machine interaction module, and at the same time, the automatic path tracking system is terminated; where the position error refers to the distance from the vehicle center of mass to the center line of the target path, and the direction error refers to the vehicle direction error relative to the target path; The vehicle stability monitoring sub-module receives the vehicle state information, estimates the vehicle center of mass side slip angle in real time, and judges whether the current estimated value of the vehicle center of mass side slip angle exceeds the preset critical value of the center of mass side slip angle; when the center of mass side slip angle exceeds the critical value, the driver is prompted to take over the vehicle and terminate the automatic path tracking system through the human-machine interaction module, and at the same time, the direct yaw moment control is started. The lower - layer execution module receives the front - and - rear wheel steering angle control laws transmitted by the steering control module, controls the front - and - rear wheel steering motors to achieve path tracking; at the same time, it receives the target longitudinal torques of the four wheels transmitted by the state monitoring module and controls the hub motors to output the target torques to achieve vehicle stability control.

7. The vehicle path tracking control system according to claim 6, wherein, The information perception module includes an on - vehicle lidar, a yaw rate sensor, a lateral acceleration sensor, and a longitudinal vehicle speed sensor; the collected state information includes: vehicle lateral position, vehicle direction angle, yaw rate, lateral acceleration, and road environment; The vehicle stability monitoring sub - module receives the vehicle yaw rate and lateral acceleration transmitted by the information perception module and estimates the vehicle sideslip angle of the center of mass in real - time.

8. The vehicle path tracking control system according to claim 7, wherein The stability monitoring sub-module receives the vehicle yaw rate transmitted by the information perception module and the lateral acceleration a y , and estimates the vehicle sideslip angle β of the vehicle center of mass in real time. According to whether the vehicle sideslip angle β of the vehicle center of mass exceeds the preset critical value β m of the vehicle center of mass sideslip angle, it is judged whether the vehicle is critically unstable. If it is critically unstable, the driver is prompted to take over the vehicle and terminate the automatic path tracking system through the human-machine interaction module. At the same time, the direct yaw moment control is started to generate an additional yaw moment ΔM according to the designed yaw moment controller, and then it is distributed and converted into the target longitudinal torques T1, T2, T3, and T4 of the wheels and transmitted to the lower-layer execution module; The vehicle sideslip angle of the center of mass is estimated using Kalman filtering technology, and the algorithm is as follows: Among them, the state value Observed value A is the state matrix, Q is the process noise covariance matrix, R is the observation noise covariance matrix, P is the estimation error covariance matrix and needs to be iteratively updated, H is the observation matrix and The yaw moment controller uses fuzzy PID control to output an additional yaw moment: where, e β = β - β d , β d = 0, K P is the proportional gain, K I is the integral gain, K D is the derivative gain, which is the sum of a predetermined value (K′ P , K′ I , K′ D ) and the fuzzy correction values (ΔK P , ΔK I , ΔK D ), i.e.: The torque distribution adopts average distribution, and the longitudinal target torque can be obtained by the following formula: Among them, T d is the driving torque, r is the wheel radius, and d is the vehicle width.

Images