ANFIS-Assisted Model Predictive Current Control for Grid-Connected PV Power Generation with Optimal MPPT Strategy

Abstract

 Conventional maximum power point tracking (MPPT) algorithms and inverter control strategies in photovoltaic (PV) power generation systems often fail to achieve optimal power extraction under environmental variations and grid fault conditions, leading to degraded dynamic performance and inaccurate tracking. To address these challenges, this paper proposes an intelligent control framework that integrates a model predictive current control (MPCC)–based MPPT strategy with an adaptive neuro-fuzzy inference system (ANFIS)–assisted MPC scheme for inverter current regulation. An identification-based PV array model is utilized to design the MPC-based MPPT controller, enabling optimal power extraction by effectively considering environmental factors such as solar irradiance and module temperature. Furthermore, the conventional finite control set model predictive current control (FCS-MPCC) is enhanced by incorporating ANFIS, which adaptively learns system behavior and assists in selecting optimal control actions. This ANFIS-MPC approach improves robustness against parameter uncertainties, nonlinearities, and external disturbances. The proposed controller predicts inverter current dynamics and determines optimal switching states to achieve precise current tracking, reduced steady-state error, and improved transient performance. The effectiveness of the MPC-based MPPT and ANFIS-assisted MPC inverter control is validated under various operating conditions using MATLAB/Simulink. Comparative analysis with conventional feedforward decoupled PI control demonstrates superior performance in terms of tracking accuracy, dynamic response, and system stability. The results confirm that the proposed ANFIS-MPC framework significantly enhances energy extraction efficiency and overall reliability of grid-connected PV systems.