Present solar photovoltaic (PV) systems incorporate a maximum power point tracking (MPPT) controller to harness the maximum power from solar arrays. The majority of the controllers perform well in the ideal conditions. However, due to the intermittent nature of the solar PVs, irradiance and temperature fluctuate rapidly, and the MPPT controller usually faces difficulties in tracking the maximum power point (MPP). In addition to this, steady-state error, oscillations during rapid variation of irradiance and temperature, and ripples in the PV voltage remain pertinent when the traditional MPPT controllers are applied. Based on recent research, an artificial intelligent method such as Adaptive Neuro-Fuzzy Inference System (ANFIS) along with a PID controller improves the tracking speed and reduces steady state error and oscillations during abrupt variation of the atmospheric conditions. However, tuning flexibility is limited to only modifying controller gains; hence lack of adaptability with ANFIS introduces computational complexity and lower transient performance. In this paper, an ANFIS-based fractional order proportional-integral (FOPI) controller is proposed to enhance the dynamic behavior of the MPPT using four Kyocera solar PV panels, configured as two panels in series and two in parallel. To validate the performance of the proposed controller, its performance is compared against a conventional incremental conductance (INC) method and an ANFIS-based proportional-integral (PI) controller. The proposed ANFIS-based FOPI controller exhibits the highest performance, as shown by the lowest root mean square error (RMSE) of 0.0475, integral time-weighted absolute error (ITAE) of 19.51, ripples in voltage of 0.38%, and transient instability of 0.01 s, which are significant when compared to previous studies. Tracking and energy extraction efficiency of the proposed controller are calculated as 99.23% and 97.7% respectively. Moreover, partial shading, uncertainty and sensitivity analysis are performed to investigate the dynamic performance of the controller under adverse conditions. The successful simulation demonstrates that the ANFIS-based FOPI controller performs better than the conventional INC controller and ANFIS PI controller.
原文链接
分享
引用
