OGPR: An Obstacle-Guided Path Refinement Approach for Mobile Robot Path Planning

摘要

Despite the emergence of the available path planning approaches for mobile robots, excessive computation time have remained an open issue, especially in time-critical scenarios. In this paper, however, an Obstacle-Guided Path Refinement (OGPR) approach is developed to plan a set of short collision-free paths between the start and the target points for mobile robots. A particle swarm optimization framework has been adopted to retrieve the obstacles geometry and subsequently refine the line-of-sight path connecting the start and the target points. The developed OGPR approach has assessed over a 2D simulation environment and the results show that its effectiveness in planning safe paths shorter than the state-of-the-art A* algorithm. This, in fact, could encourage further application of the proposed OGPR approach in future in 3D spatial environments.