Relevant Region sampling strategy with adaptive heuristic for asymptotically optimal path planning

Sampling-based planning algorithm is a powerful tool for solving problems in high-dimensional state spaces. In this article, we present novel approach to sampling the most promising regions, which significantly reduces time-consumption. The RRT# defines Relevant Region based on cost-to-come provided by optimal forward-searching tree. However, it uses cumulative cost of direct connection between current and goal as cost-to-go. To improve path efficiency, propose batch method that samples refined with strategy, defined according adaptive cost-to-go, taking advantage various sources heuristic information. proposed allows build search tree direction area, resulting superior initial solution quality reducing overall computation time compared related work. validate effectiveness our method, conducted several simulations both SE(2) SE(3) And simulation results demonstrate superiorities algorithm.