Jiandong Zhong and Jianbo Su
Robot path planning, narrow passages, probabilistic roadmap method, roadmap sampling, configuration space
Using the probabilistic roadmap method (PRM), it is difficult to generate an efficient path through narrow passages within a robot workspace due to the unreasonable number of points or milestones present in the limited space. This paper proposes a new roadmap sampling method, the randomized star builder (RSB), which efficiently identifies the narrow passages in a workspace. Furthermore, a hybrid path planning strategy based on the RSB is presented, which samples points in the corresponding configuration space. As a result, the density of milestones in a given roadmap is increased within narrow passages. Moreover, global points configured by the uniform sampler improve the milestone distribution and thus increase the path planning efficiency. Both the simulation and the experimental results demonstrate the effectiveness of the proposed algorithms.
Important Links:
Go Back
