Adaptive differential game for modular reconfigurable satellites based on neural network observer

Abstract

This paper mainly investigates an attitude coordination control method among modular reconfigurable satellites after on-orbit self reconfiguration. In the proposed problem, the satellite modules are taken as players in the differential game to make it possible for individual module to update its control strategy in real time according to its independent optimization performance requirements and the information of other modules. In addition, the adaptive dynamic programming (ADP) control is applied to design the evaluation neural network algorithm of the control strategy online learning, and a neural network observer is adopted to estimate the angular velocity. Then, an aperiodic update event-triggered control mode combined with the non-zero sum Nash strategy is proposed in order to save the communication and computing resources. Simulations are set to verify the performance of the proposed control scheme.