Lane Change Decision and Trajectory Planning for Intelligent Cars in Curved Road Scenarios

Abstract

In order to improve the efficiency of curved road passage and the safety of lane changing, the curved road section is used as the research object, and the curved road lane changing model is established by analyzing the lane changing decision and trajectory planning of intelligent vehicles. In the lane change decision, the influence of different driving styles, curved road radius, car type and other factors are considered, and the improved car ellipse model is proposed and applied to the minimum safety distance for free lane change of intelligent vehicles. In terms of trajectory planning, different lane change executable regions are proposed for drivers with different driving styles, and the optimal lane change trajectory within the lane change executable region is obtained by filtering the trajectory clusters under the Frenet coordinate system through the evaluation function. MPC control is used for trajectory tracking. Finally, the validity of the intelligent vehicle curved road lane change model is verified through Carsim and Simulink simulation.