Improved Informed RRT* Using Gridmap Skeletonization for Mobile Robot Path Planning

摘要

An improved Informed RRT* algorithm was proposed to reduce computational time, even in complex environments. Unlike RRT* used for Informed RRT* to find an initial solution for the whole configuration space, the gridmap skeletonization approach is applied to generate the initial solution. Because the skeletonization process can be performed faster and more robustly than a process where the tree is constructed by random sampling of RRT*, the initial solution is obtained efficiently and is optimized earlier compared with conventional Informed RRT* algorithms. The simulation results of the proposed improved Informed RRT* algorithm show that the computational time is shorter, with a smaller standard deviation than that of the conventional Informed RRT*.