A Multi-robot Pattern Formation Algorithm Based on Distributed Swarm Intelligence

This paper presents a solution to the problem of pattern formation on a grid map, for a homogeneous multi-robot system. In this paper we propose a natural swarm inspired algorithm based on the Particle Swarm Optimization (PSO) model and Virtual Pheromone mechanism. Basically, a virtual pheromone trail based method is proposed as the message passing mechanism among the robots, where robots make distributed movement decisions through local interactions. For one individual robot, there are two working modes, exploration and dispersion, with different indicators in the PSO model. By cooperating and communicating through the virtual pheromone, agents of the multi-robot system switch between the two phases. The PSO method helps to allocate reasonable robots to different parts of the predefined pattern. A series of experiments on simulator is carried out and proves the convergence and excellent scalability of our algorithm. By optimizing some parameter in the PSO model with the help of the simulator, the efficiency of pattern formation is further improved.