| The significance of multi-robot teams is increasing due to the growing need to coordinate distributed operations. In the future, the costs of robots will go down, and the robots will get more compact, and more capable. It is more flexible and more fault tolerant to use several simple robots than having a single powerful robot for each separate task. For a multi-robot team, an interesting topic of study is the formation control problem.; The main objective of this research is to present a novel application of hierarchical formation control for a multi-robot team. The hierarchical formation allows us to analyze the multi-robot team in a systematic way. We formulate the problem as the four-level hybrid architecture that accommodates both centralized and decentralized implementations. At the control level, we derive four control strategies---collision avoidance, position-approaching, position-keeping, and grouping---to move the distributed robots to their desired positions inside the formation. Our approach is based on the potential function method. The distinct feature of this approach is that it requires only simple computations, in particular for large numbers of interacting robots. The proposed control strategies are suitable for on-line, real-time formation tasks. Furthermore, at the formation level, we devise algorithm to form a feasible hierarchical formation. They allow a multi-robot team to generate the desired formation patterns to serve various functions. An application of five mobile robots is presented. Simulation shows the efficacy of the proposed control strategies and algorithm. |
