Real-time map building and autonomous motion planning for unmanned ground vehicles

This paper studies local elevation map building and autonomous motion planning based on feasible direction angle for unmanned ground vehicles (UGV). Founding on the configuration of sensor system mounted on unmanned ground vehicle, an algorithm of building local map based on multiple lasers is introduced, and the influence of dynamic obstacle is wiped out by combining with an algorithm of tracking dynamic obstacle. Basing on the real-time local map, feasible direction angle and vector field histogram (VFH) are presented, as so to accomplish the target of autonomous motion planning for unmanned ground vehicles in different scenes and operating states. Experiment results and data analysis prove the validity and practicability of the proposed approaches.