Remembering exploration based single-query probabilistic path planning
Jung-Tae Kim, Daijin Kim · 2010
In this paper we introduce a novel planning algorithm which we call the remembering exploration based single-query probabilistic planning algorithm, RESPP. It uses a remembering exploration algorithm as extension method of the tree data structure of RESPP. RESPP discriminates between explored nodes and unexplored nodes in the tree data structure, and uses only the unexplored nodes for the extension of the tree data structure. Less number of used nodes indicates the low computation load of the planning algorithm as shown in the experimental analysis. We also introduce its variant algorithm, bidirectional RESPP. For comparing the performance of our algorithms with other planning algorithms, we had experiments in 2D and 3D configuration space and got better performance from our suggested algorithms.