Fast H-infinty Control of Uncertain 2D Roesser Systems: A Learning-Driven LMI Approximation Approach
This paper proposes a machine learning-assisted framework that replaces computationally intensive online LMI optimization with a trained supervised model to rapidly predict state-feedback gains for delay-dependent H-infinity control of uncertain 2-D Roesser systems, achieving comparable stability and disturbance rejection with significantly improved efficiency for real-time applications.