Robust Cooperative Pursuit-Evasion Problem for Multiple Evaders and Pursuers with Measuring Noise

This paper addresses the robust cooperative pursuit-evasion problem for multiple evaders and pursuers in a bounded area with measuring noise. First, by considering the priorities of the evaders and the distances between the evaders and the pursuers, a choosing strategy is proposed for the pursuers to choose an appropriate evader as target in each period. Then, set-membership filter technique is used to deal with the measuring noise effect and the evader’s true position is estimated in an updated ellipsoid area. In particular, the ellipsoid area will reduce gradually when the pursuers come near to the evader. Finally, optimal control law is designed for each pursuer with the area-minimization policy, which guarantees that the evaders could be captured one by one in finite time. Simulation results verify the effectiveness of cooperative pursuit algorithm.