Global Path Planning for Unmanned Ground Vehicles

摘要

This paper is an overview of high-level path planning methods used in mobile robotics with special emphasis on outdoor planning for unmanned ground vehicles. It surveys all portions of the path planning process including world representation, graph search algorithms, and planning for partially and completely unknown environments. Planning representations such as Cell Decompositions, Roadmaps, and Potential Fields are covered as well as both heuristic and non-heuristic methods of graph search. Specific recently developed and popular algorithms are also investigated such as A*, D*, Potential Fields, Wavefront Planning, Probabilistic Roadmaps and Rapidly Exploring Random Trees.