Backstepping Control Method with Sliding Mode Observer for Autonomous Lane Keeping System

Abstract

This study derived a novel reduced second-order model for an autonomous lane keeping system. The proposed reduced model of the lateral vehicle motion has the following two advantages: first, one can control the vehicle's lateral motion with only simple linear second-order dynamics and second, the state variable of the reduced model includes look-ahead distance likewise human driver. The backstepping control for the lateral control and the compensation of the system parameter and uncertainties is developed using the reduced model. Moreover, the reduced model-based sliding mode observer is designed to estimate the lateral velocity. The stability of the closed-loop system is proven using passivity. The lateral control performance of the proposed method is validated via numerical simulations using CarSim and MATLAB/Simulink and compared to the fourth-order lateral motion model-based linear quadratic controller.