Abstract
This research validates An Adaptive Fuzzy Logic Controller (AFLC) has been developed for grid-connected photovoltaic (PV) systems. The primary objective of this implementation is to enhance the PV system's power generation efficiency. For achieving this, techniques of Maximum Power Point Tracking (MPPT) are utilized, which are essential to extract the highest possible power outing from PV panels. Recent developments in MPPT methods focus on improving control strategies to ensure efficient operation and smooth integration with the grid. The performance of the AFLC is extensively evaluated and compared with other controllers, like fuzzy-logic controller (FLC) and Proportional Integral (PI). The proposed AFLC controller's performance is evaluated with other methods to verify its effectiveness. To validate this method, the system is tested using MATLAB/Simulink simulations, along with experimental evaluations conducted on the control strategies are executed in real-time utilizing the DSpace DS1104 control. Experimental results show that the AFLC outperforms both the FLC and PI controllers in several key performance areas. Specifically, the AFLC demonstrates faster response times, higher convergence rates, decreased peak overshoot, minimal undershoot, and lower the error of the mean square. Additionally, the Compared to conventional Fuzzy Logic Control (FLC) and PI controllers, the AFLC delivers superior efficiency and transient response, and oscillation reduction. Compared to the FLC, the AFLC enhances tracking of power by 68.26%, and it achieves 86.25% improvement over the PI controller. These findings highlight the AFLC's potential as a highly effective and reliable optimization tool for maximizing the output power of the systems of PV. Furthermore, integral absolute error (IAE) is used as a performance metric for the PV system connected to grid to assess the efficiency of the AFLC. The AFLC demonstrated superior performance over other methods, achieving a 20% increase in PV output power compared to traditional FLC and a 30% improvement over PI controllers. The errors of the PI, FLC and AFLC approaches, each utilizing five controllers, are estimated. The error of mean square is reduced by 79.67% in comparison to PI and by 66.5% in comparison to FLC.