Optimization of Proportional-Integral Controllers of Grid-Connected Wind Energy Conversion System Using Grey Wolf Optimizer Based on Artificial Neural Network for Power Quality Improvement

This research presents a combination of artificial neural network (ANN) with the grey wolf optimizer (GWO) to improve the power quality of a grid-connected distributed power generation system (DPGS). To assess the effectiveness of the proposed algorithm, a grid-tied of small-scale wind energy conversion system (WECS) is chosen. The term power quality refers to voltage and frequency regulation, and limited harmonics. Power quality improvement is achieved through the cascaded control system's optimal tuning of three proportional-integral (PI) controllers of the grid-side inverter (GSI). However, because the DPGS model is computationally costly, the Artificial Neural Network (ANN) model is utilized as an alternative model for DPGS. Furthermore, the ANN model is employed in conjunction with the GWO to boost the optimization precision and minimize the execution time of GWO. The considered power system was repetitively simulated to obtain the input-output datasets, which validate and train the ANN model. According to the ANN model's performance evaluation, the correlation coefficient (R) is close to one, while the mean squared error (MSE) is near zero. These findings demonstrate the ANN model's great accuracy in approximating the DPGS model. Using MATLAB/Simulink, the system's performance is evaluated using the optimum values obtained using GWO-ANN for various wind speed profiles. It showed the suggested power quality method’s improved stability, convergence behavior, the effectiveness of the control mechanism, and the robustness of the proposed topology.