Dynamic Path Optimization for Robot Route Planning

Abstract

Robot is an autonomous system that integrates advances AI technologies. This paper deals with the adaptive path planning and optimization problems for robots in dynamic environments. We propose a novel route planning method based on the maze representation of workplace layouts. We generate a universal path tree by a path optimization algorithm. Then, any given entrances and exits of target nodes can be reduced to a deterministic path searching problem. Our method can quickly determine the optimal path between any pair of entrance/exit nodes. The maze-based method provides an efficient and robust route planning solution for robots in real-time and dynamic workplaces. Experiments have demonstrated the effectiveness of the method beyond traditional heuristic technologies.