Cooperative game strategy via state estimation for spacecraft with incomplete information

This paper investigates the pursuit-evasion-defense (PED) game problem for spacecraft, where the defender cooperating with the evader is to escape from the pursuer. To achieve this task, a cooperation game strategy is proposed based on the linear quadratic differential game (LQDG) and the state estimation algorithm. Considering the actual situation that the target cannot obtain complete information about the pursuer, an extended-state sliding mode observer is designed to estimate the incomplete information, including the velocity and the controller. Based on the estimated information, a control compensation game strategy is derived to weaken the impact of incomplete information on the game outcome. Furthermore, the cooperative game strategy based on the LQDG is developed, which aims at improving the escape success rate of the target under weak maneuvering conditions. The cooperative strategy can entice the pursuer to approach the defender by driving the target to maneuver behind the defender. This can increase the capture probability against the pursuer and shorten the game time. Finally, numerical simulations are performed to evaluate the proposed game strategy. The results demonstrate that the cooperative game strategy can drive the defender to intercept the pursuer while reducing energy costs and game time. Moreover, the proposed method exhibits robustness with respect to variations in orbital altitude.