A RRT*FN Based Path Replanning Algorithm
Baiming Tong, Qingbao Liu, Chaofan Dai · 2019
A path replanning algorithm based on RRT*FN(Rapidly-exploring Random Tree Fixed Nodes) is proposed for online local path planning of robot. First, we propose a procedure to reuse the tree from the last planning. Second we design a strategy to balance the exploitation of the old tree and the exploration of the current environment. Finally, the RRT*FN’s strategy is adopted to control the size of the tree. Empirical studies have shown that when the positions of the starting point, the goal and the dynamic obstacles change within a certain range, the proposed algorithm can significantly improve the quality of the solution within a limited time compared to totally starting a new planning using RRT*FN. We also compared the proposed algorithm with the two related replanning algorithms, ORRT* (Online Rapidly-exploring Random Tree*) and RT-RRT* (Real Time Rapidly-exploring Random Tree). The proposed algorithm is better with respect to the time used to find the first feasible path and the cost of the first feasible path.