Research on lane changing and overtaking for Intelligent Vehicle based on vision navigation

Median altering, Sobel operator and the Otsu algorithm were introduced to preprocess the lane image of intelligent vehicle. And the improved Hough transformation algorithm was used to extract the road characteristics and detect the lane edge. According to the prediction result of the Kalman filtering, the area of interest (AOI) of the lane edge was established and the AOI´s size can adjust dynamically to track lane edge accurately. In order to guarantee intelligent vehicle tracks expected trajectory of lane changing and overtaking steadily, controller design is one of key technologies. Establish vehicle kinematics model firstly, construct state space mathematics model tracking trajectory closed loop control system by adopted vehicle position deviation. Based on backstepping function control algorithm, choose Lyapunov function and design controller for lane changing and overtaking. Simulation and experiment result show that the design of controller for the intelligent vehicle is effective and feasible.