An efficient layered robot path planning approach based on MOPSO algorithm

Abstract

A layered path planning approach is designed to find a collision free and global optimal path from the start position to goal position in static environment. In the first two levels, called preprocessing, Delaunay triangulation and Dijkstra’s algorithm are applied to generate the initial approximate optimal paths. Then considering two objectives, minimizing the path length and maximizing the path smoothness, the improved multi-objective particle swarm optimization is used to optimize the initial paths in the third level. Various experimental results in different environments show that the proposed layered planning method can avoid the local optimal path and accelerate the convergence.