Path Planning and Smoothing in Maze Exploration Using Virtual Mobile Robot-Based Modified Probabilistic Road Map

摘要

Planning an efficient and smooth path has become a recurring issue supporting mobile robot navigation. In the case of maze exploration, path planning must provide the shortest and easiest path to be followed by a mobile robot. This paper aims to develop a path-planning method connecting several target positions with a smooth path. This paper proposes a modified probabilistic road map with virtual mobile robot path smoothing. A new approach to producing random nodes that differ from previous path planning is presented here. A virtual non-holonomic mobile robot follows the sharp turn path generated by the proposed method. The trajectory left by the virtual robot results in a smooth path connecting some desired destinations. The magnitude of the distance of the proposed path planning and smoothing was more efficient than the path of previous research, with an average value of 18.72%. The virtual mobile robot-based modified probabilistic road map successfully improved the smoothness of the path.