Improved Load-Frequency Control Using a Fuzzy Model Predictive Control Method Enhanced by Iterative Learning in Hybrid Power Systems

Abstract: This paper proposes a hierarchical hybrid control framework, termed Fuzzy Model Predictive Control with Iterative Learning (FMPC–IL), to enhance load–frequency control (LFC) in single-area hybrid power systems with high renewable penetration. The architecture integrates three layers: (i) a supervisory fuzzy logic controller (FLC) that adaptively tunes the MPC cost-function weights (Qⓜ,R)using the frequency error and its change, improving robustness under operating uncertainties; (ii) a constrained model predictive control (MPC) layer that computes the optimal control action while explicitly enforcing actuator and operational limits; and (iii) an iterative learning control (ILC) layer that exploits repetitive daily patterns in load and renewable profiles to generate a feedforward compensation signal updated cycle-to-cycle. The nonlinear governor–turbine limitations, including valve deadband, actuator saturation, and generation rate constraints (GRC), are represented in the simulation plant, whereas a linearized discrete-time model is used as the internal prediction model for MPC optimization. The proposed FMPC–IL strategy is evaluated in MATLAB/Simulink under stochastic load disturbances, renewable fluctuations, and parametric uncertainties. The results demonstrate improved frequency regulation and robustness compared with MPC, fuzzy-MPC, and MPC–IL baselines, particularly during severe disturbances and uncertainty scenarios.