Decentralized H_∞ Observer-Based Attack-Tolerant Formation Tracking Network Control of Large-Scale LEO Satellites via HJIE-Reinforced Deep Learning Approach

In this study, an H ∞ attack-tolerant decentralized observer-based formation tracking control strategy is designed for the network system (NCS) of large-scale satellite team under external disturbance and attack signal. First, smoothing signal models signals are embedded in each to avoid their corruption on state estimation Luenberger observer compensate effect satellites. addition, NCS must efficiently attenuate disturbance, measurement noise coupling from adjacent For proposed satellites, needs solve a very complicated but decoupled Hamilton Jacobi Isaacs equation (HJIE). Therefore, HJIE-reinforcement learning-based deep neural (DNN) employed directly corresponding nonlinear partial differential control-observer-coupled xmlns:xlink="http://www.w3.org/1999/xlink">HJIEi problem. When trained by Adam learning algorithm, DNN can be reinforced gain as well worst-case noise, off-line training phase. That is, algorithm-based scheme achieve robust strategy. HJIE-reinforcement-based algorithm converges, we show that reinforcement approach theoretical satellites NCS. simulation example, with wireless communication malicious given validate effectiveness method separately.