A Competition-Based Distributed Scheme for Solving Multi-Robot Cooperative Pursuit Problems

The cooperative pursuit task is a critical issue in multi-robot problem. For a dynamic task allocation in distributed multi-robot system, not only the allocation should approximate the optimal one, but also the algorithm should have high efficiency and a low computation. This paper proposes a competition-based distributed coordination control scheme for multi-robot cooperative pursuit in a dynamic environment. The proposed scheme is based on the distributed multi-target k-winner-take-all (k-WTA) algorithm, which efficiently achieves task allocation on the distributed and competitive mechanism. In the proposed approach, pursuers compete to be allocated the evaders for lower system cost. Simulative experiment results show that the proposed scheme is efficiently capable of achieving various cooperative pursuit in different scenarios.