A Bug-based Path Planner Guided with Homotopy Classes

Abstract

This paper proposes a bug-based path planning algorithm guided topologically with homotopy classes. Homotopy classes provide a topological description of how paths avoid obstacles in the workspace. They are generated with a method we developed, which builds a topological environment based on the workspace that allows to compute homotopy classes systematically. The homotopy classes are sorted according to a heuristic estimation of their lower bound. Only those with the smaller lower bound are used to guide the path planner we propose, called Homotopic Bug (HBug), which efficiently computes paths in the workspace that accomplish homotopy classes. Results show the feasibility of our method. A comparison with well-known path planners has also been included