A Fast Path Planning Algorithm Fusing PRM and P-Bi-RRT

Abstract

The path planned by the rapidly expanding random tree (RRT) algorithm is tortuous and the planning speed of RRT algorithm is slow. The path planning failure rate of the probabilistic roadmap (PRM) algorithm is high in complex scenarios. To solve these problems, this paper proposes a fusion algorithm fusing PRM and probability-based bidirectional RRT (P-Bi-RRT), which divides the planning area into two areas on average, and uses the PRM algorithm with faster planning speed for path pre-planning in each area, and then selects a node in each of the two areas to form a pair of optimal matching-points. The optimal matching-points is used to connect the two regional paths with the P-Bi-RRT algorithm or method of barrier-free direct connection. Finally, the planned path is optimized twice, the redundant nodes are trimmed and the path length is shortened. After a large number of simulation experiments, it is shown that the time of path planning is reduced by about 71% to 80%, the number of nodes is reduced by 70% to 80%, and the path length is shortened by 20% to 30% compared with the RRT algorithm.