Autonomous Vehicle Motion Planning Based on Improved RRT* Algorithm and Trajectory Optimization

An autonomous vehicle motion planning algorithm based on improved rapidly-exploring random tree (RRT) and Bezier control point optimization is proposed to accelerate the search, avoid tortuous paths and improve path smoothness. The proposed algorithm combines RRT* with the probability sampling, multi-step expansion and path simplification to generate an initial trajectory. This trajectory is used to calculate a set of initial control points of Bezier curve. Then the control points are optimized by sequential quadratic programming to improve path smoothness and safety in the environment with dynamic obstacles. The proposed algorithm is compared with the conventional method in the simulation. The results show that the algorithm can reduce the search time, improve the path smoothness and the trajectory satisfies the vehicle dynamics constraint. © 2022 Science Press. All rights reserved.