Bearing angle measurement based cooperative pursuit-evasion game in non-convex environments

This paper addresses a cooperative pursuit-evasion game in non-convex environments, for which a team of pursuers are equipped with onboard bearing angle measurement sensors. Different from many existing work, our work assumes that the agents have no global coordinate system and thus can not measure the absolute positions, but it is expected for the pursuers to estimate the positions of other agents in their local frames. Based on the estimated positions, a distributed computation method is developed where a series of Hamilton-Jacobi-Bellman (HJB) equations are solved cooperatively in order to determine the control input for every pursuer with the help of limited information exchange among pursuers. We demonstrate the effectiveness of the proposed scheme by simulations.