Robust H∞ Control of Unknown Discrete-Time Linear Systems with Time-Varying Uncertainties

• DOI: 10.1109/csis-iac60628.2023.10363864
• Publication Date: 2023-12-27T19:17:08Z
• Authors: Liao Zhu; Chunxiuzi Liu; Yu Liu; Ping Guo

The discrepancies between the mathematical model of control systems and real dynamical seriously affect performance controller. This paper considers <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$H$</tex> <inf xmlns:xlink="http://www.w3.org/1999/xlink">∞</inf> problem for unknown discrete-time linear with time-varying uncertainties. A robust controller is designed using on-policy off-policy reinforcement learning (RL) methods. Interestingly, once original system uncertainties transformed into an auxiliary a scale factor, generalized turned scaled one. As result, new algebraic Riccati equation (ARE) derived. When dynamics state are unknown, RL first introduced to solve ARE, then used learn data-driven solution ARE. It demonstrated that methods have equivalent convergence property.