Distributed Game-Theoretic Control Approach for Attitude Takeover With Microsatellites

This article presents a distributed cooperative attitude takeover control approach for failed spacecraft using multiple microsatellites, developed within a game-theoretic framework. The proposed approach integrates model predictive control with game theory, explicitly addressing reaction wheel actuator characteristics and system performance factors, including attitude error, control effort, actuator saturation and angular momentum balance. Each microsatellite optimizes its own cost function to reach a Nash equilibrium (NE), which is theoretically proven to guarantee the global optimality of the overall control problem. To overcome the limitation that each microsatellite can access only local neighbor information, an iterative approximation algorithm is developed for aggregate information estimation, relying exclusively on the inter-satellite communication topology. On this basis, a distributed NE seeking algorithm is proposed and proved to be convergent. The numerical simulation verifies the effectiveness and efficiency of the proposed algorithm under limited communication, in comparison with the centralized control method. Furthermore, the physical experiment demonstrates its robustness to disturbance and sensor uncertainties in the practical environment.