The planning of collision-free paths for a handarm\udrobotic system is a difficult issue due to the large number\udof degrees of freedom involved and the cluttered environment\udusually encountered near grasping configurations. To cope\udwith this problem, this paper presents a novel importance\udsampling method based on the use of principal component\udanalysis (PCA) to enlarge the probability of finding collisionfree\udsamples in these difficult regions of the configuration\udspace with low clearance. By using collision-free samples\udnear the goal, PCA is periodically applied in order to obtain\uda sampling volume near the goal that better covers the free\udspace, improving the efficiency of sampling-based path planning\udmethods. The approach has been tested with success on\uda hand-arm robotic system composed of a four-finger anthropomorphic\udmechanical hand (17 joints with 13 independent\uddegrees of freedom) and an industrial robot (6 independent\uddegrees of freedom).
展开▼
