Collision-Free Path Planning for Intelligent Vehicles Based on Bézier Curve

Abstract

Intelligent vehicles are expected to avoid collision through emergency lane change when a vehicle suddenly appears. Therefore, it is very meaningful to plan a path for collision-free lane change with a vehicle in front, which strikes a balance between vehicle performance and driving comfort. In this paper, a path planning method is designed based on fifth-order Bézier curve for lane change with a vehicle in front. Firstly, the boundaries were determined to ensure the fast response to the changing environment and driving comfort. Then, a multi-objective optimization function was set up, considering the curvature at the start point, the maximum curvature, the angle between path centre and road axis (the maximum swing angle), and the lane change time. Next, the coordinates of the control points that determine the curve shape were solved, and thus the optimal path was set up based on the fifth-order Bézier curve. Simulation results show that the proposed method could output suitable collision-free paths for lane change at different vehicle velocities. The research results shed new light on the dynamic path planning in actual traffic environment and enjoy great application value.

Keywords

• Intelligent vehicle
• collision avoidance
• path planning
• Bézier curve
• multi-objective optimization