Indoor Robot Path Planning Based on an Improved Probabilistic Road Map Method

Abstract

Aiming at indoor robot path planning, an improved probabilistic road map method is proposed to decide feasible routes for robots. Firstly, the proposed method sets nodes around the obstacles on the map randomly. By connecting starting point, nodes, and ending point, multi alternative practical routes which are not in collision with obstacles can be acquired. Secondly, this paper tries to optimize a route by eliminating unnecessary nodes and repositioning the nodes between starting point and ending point. Finally, the route with the shortest distance and the minimum number of inflection points is chosen as the best route. Experimental results show that, compared with other path planning methods, the improved probabilistic road map method can generate high quality routes rapidly and stably.