Locally Guided Multiple Bi-RRT∗ for Fast Path Planning in Narrow Passages
Xin Shu, Fenglei Ni, Zhou Zhou, Yechao Liu, Hong Liu, Tian Chun Zou · 2019
Rapidly-exploring Random Tree Star (RRT∗) and its variants can provide a collision-free and asymptotic optimal solution for many path planning problems. However, it is inefficient for many RRT∗ based variants to rapidly find one initial solution in a clustered environment with narrow passages, since consuming high memory as well as time, due to a large number of iterations in sampling critical nodes. To overcome this problem, the paper proposes the Locally Guided Multiple Bi-RRT∗ (LGM-BRRT∗) method, which can provide a fast solution by incorporating an improved bridge-test and a novel search strategy based on local guidance. It ensures an accelerated success rate and more efficient memory utilization compared with Bidirectional RRT∗(BRRT∗), and it is easy for implementation as well. It was verified on different types of scenarios in terms of efficiency and success rate. The results demonstrated that LGM-BRRT∗ is beneficial for fast path planning in a clustered environment with narrow passages.