Performance Evaluation of Various Adaptive Neuro Fuzzy Inference System Based Maximum Power Point Tracking for Photovoltaic System
The maximum output power of a photovoltaic (PV) system with a DC-DC converter depends mainly on the solar irradiance (G) and the temperature (T). Therefore, a maximum power point tracking (MPPT) mechanism is required to improve the overall system. The conventional MPPT approaches such as the perturbation and observation (P&O) technique have difficulty in finding true maximum power point. Thus various intelligent MPPT systems such as fuzzy logic controllers (FLC) are recently introduced. In FLC based MPPT, selecting the type of the membership function (MF) and the number of the fuzzy sets (FS) is critical for better performance. Thus, in this paper various adaptive neuro fuzzy inference system (ANFIS) is utilized to automatically tune the FLC membership functions instead of adopting the trial and error method. To find suitable MF for FLC, ANFIS is developed in MATLAB/Simulink and the effect of different types MF investigated. Simulation result shows that the FLC with triangular MF and seven FS give the best result. The evaluation indices used in this study includes the maximum extracted energy, minimum standard deviation of the error, and minimum mean square error.