Research on the Dynamic Path Planning of Manipulators Based on a Grid-Local Probability Road Map Method

A Grid-Local Probability Road Map (PRM) method was proposed for the path planning of manipulators in dynamic environments. Based on idea boundary discretization, a double-grid model built to obtain mapping from obstacles configuration space. The collision detection simplified as data indexing process improve its efficiency. Times detections were reduced by employing local programming strategies and stratified sampling method. Moreover, validity increased. Taking PUMA560 manipulator research object, simulation experiments show that time consumption collision-detection algorithm is about 14% standard one, beneficial generation probability maps compared with simple random experiment static shows consumes an average 10ms, which superior comparison has high efficiency real-time performance. 7ms per step, better than algorithm. can adjust global real avoid environment changes. mentioned been proved be efficient.