Swarm Intelligence: Swarm-based formations for groups of mobile robots

Vo Van Quang

Humans are the most social of animals: we live together in families, tribes, nations, behaving and thinking according to the rules and the norms of our communities. We can envision a future in which robots can become cooperative partners than mindless machines. In fact, some robotic tasks require an accomplishment of a group of robots such as transportation, surveillance, agriculture, defense, health-care and many others. Multiple robots in cooperation outperform independently-operating robots in terms of flexibility and fault tolerance. One of critical obstacle in the deployment of groups of mobile robots is related to their limited navigational abilities, especially when they do tasks in an unknown area. This lead to need for formation control strategies. To coordinate a group of mobile robots, swarm intelligence is adopted in order to obtain near-optimal settings of the control parameters. Here, the motion of mobile robots can be treated as an aggregation of a group of living species evolving under social interactions, e.g. bird flocks or fish schools. Having done this, the cooperative control of the robots is formulated as an optimization problem, where the particle swarm optimization (PSO) algorithm is applied to satisfy the multi-objective goal of formation establishment and collision avoidance.