A hierarchical global path planning based on multi-objective particle swarm optimization

In this study, a novel hierarchical global path planning approach for mobile robot navigation in a clutter environment is proposed. This approach has a three - level structure to obtain a feasible, optimal and safe path. In the first level, the triangular decomposition method is used to quickly establish a geometric free configuration space of the robot. In the second level, Dijkstra's algorithm is applied to find a collision - free path used as input reference for the next level. Lastly, a proposed particle swarm optimization called constrained multi-objective particle swarm optimization (CMOPSO) with an accelerated update methodology is employed to generate the global optimal path with the focus on minimizing the path length and maximizing path smoothness. The simulations illustrates the superiority of this method in terms of solution quality and actual execution time.