Autonomous Decision-Making of Path Re-planning for UGV

Abstract

A novel approach of autonomous decision-making of path re-planning for UGV is proposed when encountering the blocked road. The system includes a global route planner and a local path planner. Based on prior information in known environment, a topological map is built first to describe connectivity of roads. The global route planner does path planning or re-planning based on the topological map to generate the global route. According to the route, A* search algorithm combined with model predictive control is used for local path planning and judgment on the blocked road. The complete autonomous decision-making process includes: judgment of the local blocked road, reversing to the fork/intersection road node, the global route re-planning and detour through the blocked road. Experiments show that in known dynamic environment, the proposed approach can effectively solve the problem of path re-planning in order to improve the autonomous traffic performance for UGV.