A New Swarm Source Seeking Behavior based-on Pattern Recognition Approach

Abstract

Recent advances in swarm robots have started making it feasible to deploy large numbers of inexpensive robots for odor localization tasks. However, coordination of swarm robots to accomplish such tasks remains a challenging problem. Due to this, there are uncertainties in the environment and in the robot itself. To make an easy and efficient swarm coordination strategy for odor localization tasks, distributed algorithm-based pattern recognition combined with swarm intelligent approach has been developed. A new simple algorithm of Fuzzy-Kohonen Networks and Particle Swarm Optimization (FKN-PSO) to achieve the odor source is presented in this paper. This paper demonstrates a group of real simple robots that are under fully distributed control can successfully search, track and find a real odor plume. The results were compared between FKN-PSO and Fuzzy-PSO to analyze the performance of the swarm robots in the process of localization. It was found that the proposed approach produces a simple algorithm, and it can solve the odor localization task more efficiently than Fuzzy-PSO. Moreover, it is suitable to be implemented on real robots to localize the source of odor in a short time, and the swarm robots have the ability for keeping formation in the group without collision.