RRT path planning algorithm with enhanced sampling

Abstract

Due to the low efficiency and low path quality of the RRT algorithm, this paper introduces a sampling-enhanced RRT(SE-RRT) algorithm with four improvements to the RRT algorithm. In the path search phase, the gravitational coefficient is added to improve the efficiency of finding the feasible path without losing randomness. In the path optimization phase, turning points are selected from nodes of the original path by global rewiring. Then the sampling space is established around the turning points for local re-optimization. Finally, a near-optimal path is obtained and can be iterated to the optimal path according to the demand. In the code implementation phase, efficient path collision detection is added to prevent local paths from crossing obstacles.