Development of Path Planning Algorithm Using Probabilistic Roadmap Based on Modified Ant Colony Optimization

Abstract

In this paper, a unique combination among probabilistic roadmap, modified ant colony optimization, and third order B-spline curve has been proposed to solve path planning problems in complex and very complex environments. This proposed approach can be divided into three stages. First stage involves constructing a random roadmap depending on the environment complexity using probabilistic roadmap algorithm. Roadmap can be constructed by distributing N nodes randomly in complex and very complex static environments then pairing these nodes together according to some criteria or conditions. The constructed roadmap contains a huge number of possible random paths that may lead to connecting the start and the goal points together. Second stage includes finding path within the pre-constructed roadmap. Modified ant colony optimization has been proposed to find or to search the best path between start and goal points, where in addition to the proposed combination, ACO has been modified to increase its ability to find shorter path. Finally, the third stage uses B-spline curve to smooth and reduce the total length of the found path in the previous stage. The results of the proposed approach ensure the feasible path between start and goal points in complex and very complex environments. Also, the path is guaranteed to be short, smooth, continuous and safe.