Self-driving automobile transverse motion self-adaptive parameter control method

Abstract

The invention discloses a self-driving automobile transverse motion self-adaptive parameter control method. The method comprises the following steps: (1) acquiring vehicle motion parameters, and establishing a two-degree-of-freedom vehicle dynamics model; (2) discretizing a coefficient matrix of the vehicle dynamics model, establishing a Riccati equation, and solving a feedback matrix K through an LQR algorithm; (3) performing mean filtering and limiting on the transverse error matrix; and (4) calculating the front wheel turning angle of the vehicle according to the feedback matrix and the error matrix, and controlling the transverse movement of the vehicle through the size of the front wheel turning angle. According to the method, the feedback matrix K is solved through the discretization speed, time waste caused by repeated iteration in the solving process of the LQR algorithm is avoided, the parameter table of the speed and the value of the feedback matrix K is established in advance, and the solving speed of the LQR algorithm is increased.

Description

technical field [0001] The invention relates to an automatic driving lateral control technology, in particular to an automatic driving vehicle lateral motion adaptive parameter control method. Background technique [0002] In recent years, the autonomous driving technology of vehicles has shown a rapid development trend. Whether autonomous driving is safe, stable, and reliable is what consumers are particularly concerned about. The lateral control of autonomous driving is related to the stability of the vehicle, so it is particularly important. Existing lateral control technologies include the use of pure tracking algorithm, Stanley algorithm, LQR algorithm, and MPC algorithm. Among them, the dynamics LQR algorithm performs dynamic modeling on the vehicle, which is more in line with the vehicle's motion characteristics in medium and high-speed motion. In the lateral control effect Performed better. [0003] The existing LQR algorithm obtains the feedback matrix K by continu...

AI Technical Summary

Problems solved by technology

[0003] The existing LQR algorithm obtains the feedback matrix K by continuously solving the Riccati equation. This method affects the matrix solution speed and cannot guarantee the real-time performance of the vehicle in the lateral control tracking, thus affecting the reliability of the lateral control of automatic driving.

[0004] Patent CN202110510779.1 discloses a lateral control method, device and vehicle of an automatic driving vehicle. The optimal matrix is ​​determined according to the LQR algorithm of the linear quadratic regulator, and the steering actuator of the control vehicle is executed by multiplying the optimal matrix and the state matrix. To obtain the steering control amount, we only mentioned solving the Riccati equation to obtain the feedback matrix. Since solving the Riccati equation is a process of repeated iterations, a lot of time is wasted, and the real-time response requirements for the automatic driving of the vehicle are very high. large, so this method sacrifices the response time of automatic driving control to a certain extent

Images