A linear quadratic regulation based steer-by-wire angle following control method

Abstract

This invention relates to the field of steering system technology, specifically a steer-by-wire angle following control method based on linear quadratic regulation. The specific method is as follows: Step 1, establish a dynamic model; Step 2, transform it into a spatial state equation form; Step 3, in the LQR controller, calculate the difference between the steering wheel angular velocity and the actual angular velocity; Step 4, calculate the error matrix; Step 5, using the angle difference and angular velocity difference as inputs, construct the system's error equation; Step 6, obtain the gain matrix by iteratively solving the Riccati equation. K Step 7: Calculate the LQR torque; Step 8: Design an observer, implement a feedforward compensation strategy, and calculate the feedforward torque; Step 9: Calculate the total torque of the system in the online control execution system. Compared with existing technologies, this method not only solves for the optimal gain through LQR, avoiding repeated parameter trial and error by engineers, but also takes into account the system convergence speed and energy consumption.

Description

Technical Field

[0001] This invention relates to the field of steering system technology, specifically a steer-by-wire angle following control method based on linear quadratic adjustment. Background Technology

[0002] Traditional automotive steering systems achieve wheel steering through a mechanical transmission mechanism connected to the steering wheel. Steer-by-wire (SBW) eliminates the intermediate shaft connecting the steering wheel and the mechanical transmission mechanism, instead using control methods to control the steering actuators in response to steering wheel signals, thereby achieving wheel steering. In SBW systems, vehicle operating conditions are complex and varied in real-world scenarios, and the addition of variable gear ratios places higher demands on steering tracking performance. Therefore, a steering tracking control method with high accuracy and robustness is needed to ensure that the wheels respond accurately and quickly to steering wheel commands.

[0003] PID algorithms are widely used in steering systems, but parameter tuning issues exist in practical applications. Engineers often struggle to obtain the optimal gain during debugging, as the gain directly affects system stability and response speed. Furthermore, for small angle requests, problems arise such as failing to reach the target angle and increased response latency. Parameter gain needs to be readjusted across different systems, resulting in poor parameter universality and weak anti-interference capabilities. Summary of the Invention

[0004] To overcome the shortcomings of existing technologies, this invention provides a steer-by-wire angle following control method based on linear quadratic regulation. This method not only solves for the optimal gain through LQR, avoiding repeated parameter trial and error by engineers, but also takes into account the system convergence speed and energy consumption. The feedforward is calculated in real time by the observer, which not only enhances the system's anti-interference capability, but also helps to improve the system's response speed.

[0005] To achieve the above objectives, a steer-by-wire angle following control method based on linear quadratic regulation is designed, including an LQR controller and an observer, characterized by the following specific method:

[0006] Step 1: Establish the dynamic model of the steer-by-wire actuator;

[0007] Step 2: Transform the equations of the dynamic model in Step 1 into the form of spatial state equations;

[0008] Step 3: When designing the LQR controller, it is necessary to obtain the state information of the steering system, including the target angle and the actual angle, with the angle difference denoted as... Target angular velocity and actual angular velocity, angular velocity difference ;

[0009] Step 4, the control input of the LQR controller is: Among them, the gain matrix Error matrix ;

[0010] Step 5: Using the angle difference and angular velocity difference as inputs, construct the system error equation based on the spatial state equation in Step 2;

[0011] Step 6: Solve the Riccati equation using an iterative method to obtain the gain matrix K;

[0012] Step 7: Based on the gain matrix K, calculate the LQR torque in the design of the LQR controller. ;

[0013] Step 8: Design an observer, implement a feedforward compensation strategy, and calculate the feedforward torque. ;

[0014] Step 9: In the online control execution system, the total torque input to the system is determined by the feedforward torque. and LQR torque Composition, the total torque of the system: .

[0015] The equations of the dynamic model in step 1 are as follows: ;in, and These are the equivalent moment of inertia and damping coefficient; It is the actual turning angle; F fri F rack T mot These are motor torque, friction, and rack force. and The equivalent steering ratio of friction and rack force.

[0016] The spatial state equation in step 2 is: , ; where, matrix ,matrix ,matrix ,matrix State variables Interference resistance System input The system output is y; K1=-B / J, K2=G1 / J, K3=G2 / J, where J and B are the equivalent moment of inertia and damping coefficient, G1 and G2 are the equivalent steering ratios of friction and rack force; θ is the actual rotation angle; F fri F rack T mot These are motor torque, friction force, and rack force, respectively.

[0017] The LQR controller described is a full-state feedback controller.

[0018] In step 5, the system error equation is: , where A is the state matrix and B is the input matrix.

[0019] In step 6, the specific equation is as follows: ; where the error weight matrix The weight matrix of the control quantity .

[0020] The feedforward compensation mentioned above is achieved by compensating for steering disturbance resistance through an algorithm, which helps to improve the anti-interference ability and response speed of the time-varying system in control.

[0021] In step 7, the weight matrices Q and R are fixed in the design of the LQR controller. However, in reality, the inaccuracy of the modeling must be taken into account. The drive-by-wire system has certain simplifications and some nonlinear time-varying situations, which will reduce the tracking accuracy of the LQR controller. Therefore, a feedforward compensation strategy is added.

[0022] In step 8, when designing the observer, the state-space equations and output quantities of the steering system need to be obtained. These were already obtained in step 2, and the state-space equations are as follows: , Among them, the system state variables of the observer Observer gain matrix K1 = -B / J, K2 = G1 / J, K3 = G2 / J, where, and These are the equivalent moment of inertia and damping coefficient. and The equivalent steering ratio of frictional force and rack force; and The observer gain coefficient, For the estimated interference resistance, For the estimated actual turning angle, The actual state value measured by the sensor. These are the estimated state values.

[0023] Compared with the prior art, this invention provides a steer-by-wire angle following control method based on linear quadratic regulation. It can not only solve for the optimal gain through LQR, avoiding repeated parameter trial and error by engineers, but also take into account the system convergence speed and system energy consumption. The feedforward is calculated in real time by the observer, which not only enhances the system's anti-interference ability, but also helps to improve the system response speed. Attached Figure Description

[0024] Figure 1 The flowchart is for the invention.

[0025] Figure 2 This is a control principle diagram of the present invention.

[0026] Figure 3 This is a diagram showing the actual vehicle slope signal response results of the present invention.

[0027] Figure 4 The diagram shows the actual vehicle sinusoidal signal response results of this invention. Detailed Implementation

[0028] The present invention will now be further described with reference to the accompanying drawings.

[0029] like Figures 1 to 4 As shown, this invention provides an angle following control method based on a feedforward linear quadratic regulator (LQR), the specific operation of which is as follows:

[0030] Step 1: Establish the dynamic model of the steer-by-wire actuator: In the formula, and These are the equivalent moment of inertia and damping coefficient. It is the actual turning angle, F fri F rack T mot These are motor torque, friction, and rack force. and The equivalent steering ratio of friction and rack force.

[0031] Step 2: Rewrite the dynamic equations from Step 1 in the form of their spatial state equations, as follows: , ; where the matrix ,matrix ,matrix ,matrix State variables Interference resistance System input The system output is y; K1=-B / J, K2=G1 / J, K3=G2 / J, where J and B are the equivalent moment of inertia and damping coefficient, G1 and G2 are the equivalent steering ratios of friction and rack force; θ is the actual rotation angle; F fri F rack T mot These are motor torque, friction force, and rack force, respectively.

[0032] Step 3: When designing the LQR controller, it is necessary to obtain the state information of the steering system, including the target angle and the actual angle, and the angle difference is denoted as... Target angular velocity and actual angular velocity, angular velocity difference .

[0033] Step 4: The LQR controller design is a full-state feedback controller, and its control quantity design is as follows: , where the gain matrix Error matrix .

[0034] Step 5: Using the angle difference and angular velocity difference as inputs, construct the system's error equation based on the spatial equations in Step 2: , where A is the state matrix and B is the input matrix.

[0035] Step 6: The purpose of the LQR controller is to find a set of system state variables when the control variable is sufficiently small, by introducing a cost function, and then solving the Ricati equation using an iterative method to obtain the gain matrix K, as follows: ; where the error weight matrix The weight matrix of the control quantity .

[0036] Step 7: In the design of the LQR controller, the weight matrices Q and R are fixed. However, in reality, the inaccuracy of the modeling must be taken into account. The drive-by-wire system has certain simplifications and some nonlinear time-varying situations, which will reduce the tracking accuracy of the LQR controller. Therefore, a feedforward compensation strategy is added.

[0037] Step 8: The idea of ​​feedforward compensation is to compensate for the interference resistance through algorithm. In this way, it is beneficial to improve the anti-interference ability and response speed of the time-varying system in control. Since the actual interference resistance is difficult to measure with sensors, this paper introduces the observer algorithm, which is used to supplement the information of the sensors.

[0038] Step 9: When designing the observer, the state-space equations and output quantities of the steering system need to be obtained. These were obtained in step 2. The state-space equations are as follows: , Among them, the system state variables of the observer Observer gain matrix K1 = -B / J, K2 = G1 / J, K3 = G2 / J, where, and These are the equivalent moment of inertia and damping coefficient. and The equivalent steering ratio of frictional force and rack force; and The observer gain coefficient, For the estimated interference resistance, For the estimated actual turning angle, The actual state value measured by the sensor. These are the estimated state values.

[0039] Step 10: In the online control execution system, the system input consists of a feedforward part and an LQR control part. The system control quantities are as follows: Among them, the LQR calculation part Feedforward calculation section .

[0040] For example, 3, Figure 4 As described above, through the settings in steps 1 to 10, the present invention not only reduces the debugging difficulty for engineers, but also improves the system's response speed and anti-interference ability, while avoiding the problem of not being able to reach the target turning angle at small angles, thus demonstrating that the present invention is more suitable for high-precision steer-by-wire systems.

Claims

1. A steer-by-wire angle following control method based on linear quadratic regulation, comprising an LQR controller and an observer, characterized in that: The specific method is as follows: Step 1: Establish the dynamic model of the steer-by-wire actuator; Step 2: Transform the equations of the dynamic model in Step 1 into the form of spatial state equations; Step 3: When designing the LQR controller, it is necessary to obtain the state information of the steering system, including the steering wheel angle and the actual angle, with the angle difference denoted as... The difference between the steering wheel angular velocity and the actual angular velocity. ; Step 4, the control input of the LQR controller is: Among them, the gain matrix Error matrix ; Step 5: Using the angle difference and angular velocity difference as inputs, construct the system error equation based on the spatial state equation in Step 2; Step 6: Solve the Riccati equation using an iterative method to obtain the gain matrix K; Step 7: Based on the gain matrix K, calculate the LQR torque in the design of the LQR controller. ; Step 8: Design an observer, implement a feedforward compensation strategy, and calculate the feedforward torque. ; Step 9: In the online control execution system, the total torque input to the system is determined by the feedforward torque. and LQR torque Composition, the total torque of the system: .

2. The steer-by-wire angle following control method based on linear quadratic regulation according to claim 1, characterized in that: The equations of the dynamic model in step 1 are as follows: ;in, and These are the equivalent moment of inertia and damping coefficient.

3. The steer-by-wire angle following control method based on linear quadratic adjustment according to claim 1, characterized in that: The spatial state equation in step 2 is: , ; where the matrix ,matrix ,matrix ,matrix State variables Interference resistance System input The system output quantity y; It is the actual turning angle; F fri F rack F mot These are motor torque, friction force, and rack force, respectively.

4. The steer-by-wire angle following control method based on linear quadratic adjustment according to claim 1, characterized in that: The LQR controller described is a full-state feedback controller.

5. The steer-by-wire angle following control method based on linear quadratic regulation according to claim 1, characterized in that: In step 5, the system error equation is: .

6. The steer-by-wire angle following control method based on linear quadratic adjustment according to claim 1, characterized in that: In step 6, the specific equation is as follows: ; where the error weight matrix The weight matrix of the control quantity .

7. The steer-by-wire angle following control method based on linear quadratic regulation according to claim 1, characterized in that: The feedforward compensation mentioned above is achieved by compensating for steering disturbance resistance through an algorithm, which helps to improve the anti-interference ability and response speed of the time-varying system in control.

8. The steer-by-wire angle following control method based on linear quadratic regulation according to claim 1, characterized in that: In step 7, the weight matrices Q and R are fixed in the design of the LQR controller. However, in reality, the inaccuracy of the modeling must be taken into account. The drive-by-wire system has certain simplifications and some nonlinear time-varying situations, which will reduce the tracking accuracy of the LQR controller. Therefore, a feedforward compensation strategy is added.

9. The steer-by-wire angle following control method based on linear quadratic regulation according to claim 1, characterized in that: In step 8, when designing the observer, it is necessary to obtain the state-space equations and output quantities of the steering system. These have already been obtained in step 2, and the state-space equations are as follows: , Among them, the system state variables of the observer Observer gain matrix .

Images