Coordenação de times de robôs baseada em mapas descentralizados de feromônio repulsivo e regras locais de autômatos celulares

Abstract

This work investigates a bioinspired model for swarm robotics coordination, focusing on the surveillance task and environmental exploration. Among the bioinspired strategies employed in the proposed model, we highlight the inverse ant system and the cellular automata. The decision related to the robot’s next move is made with the aid of the repulsive pheromone, which allows the search for regions that have not been recently visited, and the modelling by cellular automata rules, that allows a local decision based on the neighbourhood information for each robot. The model presents a new repulsive pheromone modelling, which enables the decentralization of navigation decisions. In this model, each robot maintains an independent pheromone map, which is continuously updated with the robot’s move itself, both in the deposit and in the evaporation of the pheromone. In addition, the individual pheromone map is eventually updated by exchanging messages with other robots that are exploring in nearby regions. Thus, individual and independent maps replace the need for using an intelligent environment that distributes the pheromone information, which is not always feasible. Besides, it eliminates the demand for a centralizing agent that stores a global map and needs to communicate with the entire swarm constantly. The proposed model was evaluated through an agent simulation software, implemented in the C programming language by the author, and in the Webots robotics simulation platform. Experiments were carried out to validate the proposed model in different environments, with different shapes and sizes, as well as varying the number of robots in the swarm. The analysis of the results showed that the model was able to perform the decentralized coordination of the team, with a good performance of the robots when executing the surveillance task.