Vehicle chassis system control method

Abstract

The invention provides a vehicle chassis system control method, which belongs to the technical field of vehicle control, and comprises the following steps: upper-layer control: respectively taking the safety, comfort and stability of the whole vehicle as three optimization targets, setting the weight of each optimization target according to environment information sensed by a vehicle-mounted sensor, realizing game balance solution, and setting the weight of each optimization target; obtaining a control target of each subsystem; middle-layer control: taking the control target of each subsystem, the environment information and the parameter information of the vehicle model as the input of machine learning, and obtaining the value of the control variable of each subsystem after the machine learning; in the lower-layer control, the values of the control variables of all the subsystems are transmitted into model prediction, the model prediction makes a control decision according to the values of all the control variables and the current state of the vehicle, control over a chassis system is achieved, and the control execution condition is fed back to the middle-layer control. According to the invention, conflicts among the systems can be effectively avoided, and the comfort, safety and stability of the vehicle can be ensured.

Description

Technical Field

[0001] The invention belongs to the technical field of vehicle control, and in particular relates to a vehicle chassis system control method. Background Art

[0002] With the rapid development of the automotive industry, vehicle chassis systems have gradually integrated more active control components and wire-controlled chassis components. However, as the number of actuators increases, the complexity of the system also increases. When the number of actuators exceeds the motion state that needs to be controlled, an overdriven system is formed. Overdriven systems allow multiple control modules to affect the same physical variables. Although this provides greater control flexibility, it also brings the risk of conflict between subsystems, which may damage the stability of the vehicle and threaten the safety of passengers. In the current chassis integrated control methods, although the decentralized architecture is simple to implement, it does not make full use of shared information, while the centralized architecture provides global optimal control performance but lacks flexibility and fail-safe redundancy. The multi-layer architecture attempts to balance these advantages and disadvantages and optimize the performance of the entire vehicle through a hierarchical control method. However, there are still problems such as high coordination complexity and limited real-time control capabilities, which are still insufficient for overall enhancement of the vehicle's dynamic performance while improving driving safety, stability, and comfort. Summary of the invention

[0003] The purpose of the present invention is to provide a vehicle chassis system control method, which can effectively avoid conflicts between systems while realizing integrated control of various chassis subsystems and can ensure the comfort, safety and stability of the vehicle.

[0004] The present invention is achieved through the following technical solutions:

[0005] A vehicle chassis system control method, wherein the chassis system includes multiple subsystems, including an upper layer control, a middle layer control and a lower layer control, specifically includes the following steps:

[0006] Step S1, upper control: taking the overall safety, comfort and stability of the vehicle as three optimization goals respectively, setting the weight of each optimization goal according to the environmental information sensed by the on-board sensor, achieving the game balance solution, and obtaining the control target of each subsystem;

[0007] Step S2, middle-level control: taking the control target of each subsystem, environmental information and parameter information of the vehicle model as the input of machine learning, and obtaining the value of the control variable of each subsystem after machine learning;

[0008] Step S3, lower-level control: the values ​​of the control variables of each subsystem are passed into the model prediction. The model prediction makes control decisions based on the values ​​of each control variable and the current state of the vehicle to achieve control of the chassis system, and feeds back the control execution status to the middle-level control.

[0009] Furthermore, in step S1, the control target of each subsystem is the intuitive effect after each subsystem executes the control command.

[0010] Furthermore, in step S1, the subsystems include a steering system, a braking system, a suspension system and a drive system, the control targets of the steering system include a steering angle and a steering acceleration, the braking system includes a braking torque and a braking acceleration, the suspension system includes a suspension stiffness and a suspension damping, and the drive system includes a drive torque and a drive acceleration.

[0011] Furthermore, in step S2, the machine learning includes a game layer and a machine learning network layer. The control objectives of each subsystem, environmental information and parameter information of the vehicle model are used as inputs of the game layer. The game layer outputs the control strategies of each subsystem. The control strategies of each subsystem and the feedback of the execution of model predictive control are used as data sets. The machine learning network layer analyzes and trains the data sets to obtain the values ​​of the control variables of each subsystem.

[0012] Furthermore, in step S3, the model predicts the execution of control of the vehicle's lateral dynamics, longitudinal dynamics and vertical dynamics according to the control decision.

[0013] Furthermore, in step S1, the security is determined by the formula The comfort is expressed by the formula The stability is expressed by the formula Indicates that, v x represents the minimum vehicle speed, d rel is a function of the relative distance of the obstacle ahead, f(d rel , v x ) is the collision risk assessment function related to distance and speed, is the square of the vehicle body yaw angular velocity, α1, α2, α3 are coefficients, and their value range is [0,1]. is the rate of change of longitudinal acceleration, is the rate of change of lateral acceleration, RMS(a z ) is the root mean square value of the vertical vibration, β1, β2, and β3 are coefficients, and their value ranges are all [0,1]. is the yaw rate error, is the roll angle change rate, and α is the tire slip angle.

[0014] Furthermore, in step S1, the overall safety, comfort and stability of the vehicle are respectively used as three optimization objectives. According to the environmental information perceived by the vehicle-mounted sensors, the weight of each optimization objective is set, the utility function is established, and the constraint conditions are set to achieve a game equilibrium solution.

[0015] The present invention has the following beneficial effects:

[0016] 1. The present invention firstly takes the overall safety, comfort and stability of the vehicle as three optimization objectives in the upper-level control, sets the weight of each optimization objective according to the environmental information perceived by the vehicle-mounted sensor, realizes the game equilibrium solution, and obtains the control objective of each subsystem. Then, in the middle-level control, the control objective of each subsystem, the environmental information and the parameter information of the vehicle model are used as the input of machine learning, and the value of the control variable of each subsystem is obtained after machine learning. Finally, in the lower-level control, the value of the control variable of each subsystem is passed into the model prediction, and the model prediction executes the control command to realize the control of the chassis system, and the control execution status is fed back to the middle-level control, thereby realizing the combination of game theory, machine learning and model predictive control to jointly form the control of the vehicle chassis system. It can not only effectively coordinate the subsystems and reduce the conflicts between the subsystems, but also avoid the influence of weak real-time control capability compared with the existing control method, which not only improves the safety and comfort of passengers, but also increases the stability of the vehicle. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] The present invention is further described in detail below with reference to the accompanying drawings.

[0018] Figure 1 It is a flow chart of the present invention.

[0019] Figure 2 It is a schematic diagram of the principle of the present invention. DETAILED DESCRIPTION

[0020] like Figure 1 and Figure 2 As shown in the figure, the vehicle chassis system control method includes four subsystems, namely, the steering system, the braking system, the suspension system and the drive system. The control target of each subsystem is the intuitive effect after each subsystem executes the control command, that is, the control target of the steering system includes the steering angle and the steering acceleration, the braking system includes the braking torque and the braking acceleration, the suspension system includes the suspension stiffness and the suspension damping, and the drive system includes the driving torque and the driving acceleration.

[0021] The control method includes upper-level control, middle-level control and lower-level control, and specifically includes the following steps:

[0022] Step S1, upper control: taking the overall safety, comfort and stability of the vehicle as three optimization goals respectively, setting the weight of each optimization goal according to the environmental information sensed by the on-board sensor, achieving the game balance solution, and obtaining the control target of each subsystem;

[0023] Specifically, the safety is given by the formula The comfort is expressed by the formula The stability is expressed by the formula Indicates that, v x represents the minimum vehicle speed, d rel =max(d front , 0) is the function of the relative distance of the obstacle in front, d front is the actual distance from the current vehicle to the preceding vehicle, is the collision risk assessment function related to distance and speed, α′ and β′ are weight coefficients, is the square of the vehicle body yaw angular velocity, α1, α2, α3 are coefficients, and their value range is [0,1]. is the rate of change of longitudinal acceleration, is the rate of change of lateral acceleration, RMS(a z ) is the root mean square value of the vertical vibration, β1, β2, and β3 are coefficients, and their value ranges are all [0,1]. is the yaw rate error, is the roll angle change rate, and α is the tire slip angle.

[0024] Set the weights of each optimization objective and establish the utility function U = w s U1+w c U2+w t U3, setting includes sideslip angle constraint |α|≤α max , distance constraint d rel ≥d min , Steering Angle Constraint Driving force constraint |T|≤T max Constraints of the game are solved using Get the control target of each subsystem, where w s 、w c 、w t are the weights of safety, comfort and stability respectively, and w s +w c +w t = 1, based on the environmental information perceived by the vehicle sensors, the emphasis on safety, comfort, and stability in a specific environment is determined, thereby determining the weight values, x * is the decision change, x i is the decision variable of the ith control objective, x -i are decision variables for other control objectives, α max d min , T max The thresholds are set respectively.

[0025] On-board sensors include lidar installed on the top of the vehicle, on the front bumper or under the window to detect obstacles, radar installed on the bumper and side body to measure distance and speed, GPS installed on the roof for positioning, inertial measurement unit installed in the center of the vehicle to measure vehicle acceleration, angular velocity and direction, and wheel speed sensors installed on the wheels to detect wheel rotation speed.

[0026] Step S2, middle-level control: taking the control target of each subsystem, environmental information and parameter information of the vehicle model as the input of machine learning, and obtaining the value of the control variable of each subsystem after machine learning;

[0027] Specifically, machine learning is a two-layer multimodal machine learning architecture, which includes a game layer and a machine learning network layer. The control objectives of each subsystem, environmental information, and parameter information of the vehicle model are used as inputs of the game layer. The game layer outputs the control strategies of each subsystem through a convolutional network. The control strategies of each subsystem and the feedback of the execution of model predictive control are used as data sets. The machine learning network layer analyzes and trains the data sets to obtain the values ​​of the control variables of each subsystem and realize the hot start of the model predictive control, thereby reducing the number of iterations of the model predictive control and speeding up the calculation time.

[0028] The vehicle model may be a 7-degree-of-freedom vehicle model, a 14-degree-of-freedom vehicle model, or a vehicle model with higher degrees of freedom.

[0029] The analysis and training of the data set includes checking whether there are missing values, abnormal values ​​or duplicate values ​​in the data set and processing them; processing the data features, such as using convolutional neural networks for normalization; dividing the data set into training sets, validation sets and test sets. According to the actual situation, select a suitable neural network model for training, and then verify and fine-tune it. After the training is completed, the predicted values ​​of the control variables of each subsystem are obtained. Among them, the neural network model can be a convolutional neural network or a recurrent neural network.

[0030] Step S3, lower-level control: the values ​​of the control variables of each subsystem are passed into the model prediction. The model prediction makes control decisions based on the values ​​of each control variable and the current state of the vehicle to control the chassis system, and feeds back the control execution status to the middle-level control;

[0031] Specifically, after the machine learning network layer training is completed, the model predictive control (MPC) is initialized, and then the model predictive control is started, and performance testing and model updating are performed.

[0032] The model predicts the execution of control decisions for the lateral, longitudinal and vertical dynamics of the vehicle, i.e., the lateral dynamics include the steering system, the longitudinal dynamics include the suspension system, and the longitudinal dynamics include the braking system and the drive system.

[0033] The above description is only a preferred embodiment of the present invention, and therefore cannot be used to limit the scope of implementation of the present invention. That is, equivalent changes and modifications made according to the scope of the patent application of the present invention and the contents of the specification should still fall within the scope covered by the patent of the present invention.

Claims

1. A vehicle chassis system control method, the chassis system includes multiple subsystems, characterized in that: It includes upper-level control, middle-level control and lower-level control, and specifically includes the following steps: Step S1, upper control: taking the overall safety, comfort and stability of the vehicle as three optimization goals respectively, setting the weight of each optimization goal according to the environmental information sensed by the on-board sensor, achieving the game balance solution, and obtaining the control target of each subsystem; Step S2, middle-level control: taking the control target of each subsystem, environmental information and parameter information of the vehicle model as the input of machine learning, and obtaining the value of the control variable of each subsystem after machine learning; Step S3, lower-level control: the values ​​of the control variables of each subsystem are passed into the model prediction. The model prediction makes control decisions based on the values ​​of each control variable and the current state of the vehicle to achieve control of the chassis system, and feeds back the control execution status to the middle-level control.

2. A vehicle chassis system control method according to claim 1, characterized in that: In step S1, the control target of each subsystem is the visual effect after each subsystem executes the control command.

3. A vehicle chassis system control method according to claim 2, characterized in that: In step S1, the subsystems include a steering system, a braking system, a suspension system and a drive system. The control targets of the steering system include a steering angle and a steering acceleration. The braking system includes a braking torque and a braking acceleration. The suspension system includes a suspension stiffness and a suspension damping. The drive system includes a drive torque and a drive acceleration.

4. A vehicle chassis system control method according to claim 1, 2 or 3, characterized in that: In step S2, the machine learning includes a game layer and a machine learning network layer. The control objectives, environmental information and parameter information of the vehicle model of each subsystem are used as inputs of the game layer. The game layer outputs the control strategy of each subsystem. The control strategy of each subsystem and the feedback of the execution of model predictive control are used as data sets. The machine learning network layer analyzes and trains the data sets to obtain the values ​​of the control variables of each subsystem.

5. A vehicle chassis system control method according to claim 1, 2 or 3, characterized in that: In step S3, the model predicts the control of the vehicle's lateral dynamics, longitudinal dynamics and vertical dynamics according to the control decision.

6. A vehicle chassis system control method according to claim 1, 2 or 3, characterized in that: In step S1, the security is determined by the formula The comfort is expressed by the formula The stability is expressed by the formula Indicates that, v x represents the minimum vehicle speed, d rel is a function of the relative distance of the obstacle ahead, f(d rel , v x ) is the collision risk assessment function related to distance and speed, is the square of the vehicle body yaw angular velocity, α1, α2, α3 are coefficients, and their value range is [0,1]. is the rate of change of longitudinal acceleration, is the rate of change of lateral acceleration, RMS(a z ) is the root mean square value of the vertical vibration, β1, β2, and β3 are coefficients, and their value ranges are all [0,1]. is the yaw rate error, is the roll angle change rate, and α is the tire slip angle.

7. A vehicle chassis system control method according to claim 1, 2 or 3, characterized in that: In step S1, the overall safety, comfort and stability of the vehicle are respectively used as three optimization objectives. According to the environmental information sensed by the vehicle-mounted sensors, the weight of each optimization objective is set, the utility function is established, and the constraint conditions are set to achieve a game balance solution.

Images