Genetic Algorithm Based Path Planning and Optimization for Autonomous Mobile Robots with Morphological Preprocessing

This paper presents an algorithm for optimal path planning for mobile robots using genetic algorithms coupled with morphological image preprocessing of the terrain. Path planning in a given environment, being a NP-hard problem, is computationally demanding especially if exact or deterministic techniques are employed. This paves the way for the use of evolutionary computing techniques, such as genetic algorithms (GAs), for the solution of this problem. However as GAs tend to find optimal paths without considering the location and distribution of obstacles in the given map, therefore a large amount of time may be spent before even a feasible path is found. The technique conferred in this paper overcomes this shortcoming by using specialized morphological preprocessing techniques applied on the landscape representative image. The results obtained on a variety of structured, unstructured and clustered terrains indicate the efficacy of the algorithm presented and the potential for the use of this algorithm in real-time operation of mobile robots