Affiliations: LRI/GReLearn, Department of Computer Science, University Badji Mokhtar, Annaba, Algeria
Corresponding author: Abdelmadjid Benmachiche, LRI (Laboratory of Research on Computing ``Informatique in french'')/ GReLearn (Group of Research in E-learning), Department of Computer Science, University Badji Mokhtar, Annaba, Algeria. E-mail:[email protected]
Abstract: In this paper, we discuss cooperative approach to solve the problem of navigation for autonomous mobile robots in fully dynamic environments, it is about building an environment of robotics specific to a construction site where such robots must cooperate to deal with the functions related to the site. For that, we used the power of the biological inspiration; the genetic algorithms (GA's) because theirs capacity to solve hard problems in short time in any environment with increasing complexity which are an interesting alternative to conventional methods of path planning. Our work is to determine the optimal path of several mobile robots using genetic algorithms in a dynamic environment. The evaluation of paths is a ``fitness'' function based on the path length. This method is implemented and tested on several scenarios. The results demonstrate the robustness and performance of our approach.
Keywords: Construction site, evolutionary robotics, genetic algorithm, mobile robots, multi-robot system