Adaptively combining multiple sampling strategies for probabilistic roadmap planning

Abstract

Several sophisticated sampling strategies have been proposed recently to address the narrow passage problem for probabilistic roadmap (PRM) planning. They all have unique strengths and weaknesses in different environments, but in general, none seems sufficient on its own. In this paper, we present a new approach that adaptively combines multiple sampling strategies for PRM planning. Using this approach, we describe an adaptive hybrid sampling (AHS) strategy using two component samplers: the bridge test, a specialized sampler for narrow passages, and the uniform sampler. We tested the AHS strategy on robots with two to eight degrees of freedom. These preliminary tests show that the AHS strategy achieves consistently good performance, compared with fixed-weight hybrid sampling strategies.