Path-planning algorithms for self-driving vehicles based on improved RRT-Connect
Jin Li, Chao‐Wei Huang, Minqiang Pan
Abstract
Abstract This study aims to solve path planning of intelligent vehicles in self-driving. In this study, an improved path-planning method combining constraints of the environment and vehicle is proposed. The algorithm designs a reasonable path cost function, then uses a heuristic guided search strategy to improve the speed and quality of path planning, and finally generates smooth and continuous curvature paths based on the path post-processing method focusing on the requirements of path smoothness. A simulation test shows that compared with the basic rapidly-exploring random tree (RRT), RRT-Connect and RRT* algorithms, the path length of the proposed algorithm can be reduced by 19.7%, 29.3% and 1% respectively, and the maximum planned path curvature of the proposed algorithm is 0.0796 m-1 and 0.1512 m-1 respectively, under the condition of a small amount of planning time. The algorithm can plan the more suitable driving path for intelligent vehicles in a complex environment.