Non-uniform Sampling-Based Path Planning Algorithm for Unmanned Vehicles in Unstructured Environments

Aiming at the problems of slow rates of convergence, low node utilization, and high randomness of the RRT∗ algorithm, this paper proposes an improved path planning algorithm integrating artificial potential field and Informed RRT*. Firstly, a non-uniform sampling approach based on the artificial potential field is proposed to reach the target point faster and avoid obstacles more effectively. Secondly, a method to adaptively adjust the step size based on obstacle distances is proposed. Finally, cubic B-spline curve are employed to achieve smooth curves. Experiments show that compared to the RRT∗ and Informed RRT∗ algorithms, our proposed algorithm exhibits shorter initial solution time, faster rates of convergence, higher efficiency, and significantly improved performance.