NT-ARS-RRT: A novel non-threshold adaptive region sampling RRT algorithm for path planning

Abstract

Rapidly-Exploring Random Tree (RRT) algorithm is a widely used path planning method. However, it suffers from low solution efficiency, no search guidance, poor quality of the obtained path and the problem of obvious reduction of search efficiency in the narrow exit environment, which greatly reduces its performance. To overcome these issues, this paper proposes a novel non-threshold adaptive region sampling RRT (NT-ARS-RRT) algorithm for path planning. First, a map preprocessing method is proposed to reduce the explore space of the random tree and improves the path search efficiency especially at narrow exits. Second, a branch-and-leaf backtracking method is proposed to optimize the selection of the nearest node. Third, a distance threshold connection constraints optimization method between the newest node and the target point is proposed to reduce the generation of unnecessary nodes and thus further improves the algorithm’s speed. Fourth, the initial path is optimized by using a rewiring method based on triangular inequality to improve the path quality. Experimental results show the effectiveness of the proposed algorithm.