Use of Particle Swarm Optimization in path planning for rough terrain

Abstract

With the advancement of autonomous vehicles the problem of navigation in rough terrain is becoming an interesting problem to tackle. One possible solution that has already been demonstrated for navigation on flat surfaces has been the use of Particle Swarm Optimization, PSO. This document aims to present results obtained from experiments gauging the adequacy of PSO for path planning in rough terrain environments. The PSO algorithm is set up to choose paths composed of straight line segments. A Controller is then used in a simulator to guide our vehicle along each path and the quality of the route traversed is then used to evaluate each path. Two different environments are prepared and solved for and the result for each environment is presented. It is shown that PSO is indeed applicable for use in path planning for rough environments.