Cellular Automata based Decentralized Cooperative Collision Avoidance Control for Multiple Mobile Robots

Abstract

One of the main challenges in the decentralized implementation of mobile multi-robotic systems is the avoidance of collisions among agents and other obstacles in cluttered environments as well as the convergence of each robot to its desired destination. This paper presents a novel decentralized, cooperative navigation algorithm for a team of mobile robots based on the concept of cellular automata that is proven to guarantee collision avoidance at all times even under the presence of static non-cooperative obstacles. The algorithm does not need to differentiate among agents and obstacles and use a time-varying localization-based priority rule to guarantee safe motion while reducing the occurrence of deadlocks. A Monte Carlo simulation is performed showcasing the performance of the algorithm under different densities of obstacles and agents.