Path planning based on adaptive space expansion-RRT* for autonomous vehicle
Yan Wang, Fei Liu, Chenyu Wang, Aidie Zheng, Haojie Huang · Transactions of the Institute of Measurement and Control · 2025
Autonomous driving systems require path algorithms that optimize concurrently real-time performance, path quality, and safety. Although the Rapid Exploration Random Tree Star (RRT*) algorithm advances path planning, its practical application is still hampered by slow convergence in cluttered environments, suboptimal path smoothing, and neglect of actual vehicle characteristics. To address these limitations, an adaptive space expansion-based RRT* algorithm (ASE-RRT*) is proposed. The algorithm improves search efficiency by gradually expanding the sampling space and employs a dynamic step-size strategy to adapt to the exploration and exploitation needs at different stages of the random tree growth process. In the study of vehicle safety and comfort, the actual shapes of vehicles and obstacles as well as the kinematic properties are fully considered, and Non-Uniform Rational B-Spline (NURBS) curves are used for optimization in combination with the path curve characteristics.