Decentralized Fuzzy $H_\infty$-Iterative Learning LFC With Time-Varying Communication Delays and Parametric Uncertainties

Load frequency control (LFC) continues to be a major problem in multi-area power systems, which is compounded by communication network issues and parametric uncertainties, that degrade controller performance. In this contribution, a fuzzy <inline-formula><tex-math notation="LaTeX">$H_\infty$</tex-math></inline-formula>-iterative learning controller (FILC) is designed for decentralized LFC with little knowledge of the local power area's model and no knowledge of the external power areas’ models. Further to this, time-varying communication delays, parametric variations, large disturbances, and non-identical power system area parameters are considered. The FILC strategy comprises a fuzzy strategy that quickly rejects large disturbances and drives the error to a designed tolerable band where an iterative learning control technique that is proven to be asymptotically stable with a prescribed <inline-formula><tex-math notation="LaTeX">$H_\infty$</tex-math></inline-formula> performance achieves zero-error convergence. The proposed FILC is compared with that of the well-known PI algorithm for a three-area power system. Results indicate that the FILC performs significantly better than the PI controller under practical combinations of network problems, overlapping large disturbances in multiple areas and parameter variations.