A path planning algorithm based on parallel particle swarm optimization

Abstract

A novel path planning algorithm based on Parallel Particle Swarm Optimization (PSO) is proposed in this paper to solve the real-time path planning problem in dynamic multi-agent environment. This paper first describes the advantages of PSO algorithm in real time search problems, i.e. path finding problems. Then considering the development trend of multiprocessors, the parallel PSO (PPSO) was proposed to speed up the search process. Due to the above mentioned advantages, we in this paper adopt the PPSO to distribute particles onto different processors. By exchanging data upon the shared memory, these processors collaborate to work out optimal paths in complicated environment. The resulting simulation experiments show that when compared with traditional PSO, using PPSO could considerably reduce the searching time of path finding in multi-agent environment. © 2014 Springer International Publishing Switzerland.