Output Voltage Stabilization of a PMSG Wind Turbine by Nonlinear ANN Identification- Based MPC

Abstract This paper presents an advanced control strategy for stabilizing the output voltage of Permanent Magnet Synchronous Generator (PMSG)-based wind turbines under variable wind speed conditions. The proposed method integrates a Model Predictive Control (MPC) scheme with a nonlinear system identification approach using Artificial Neural Network (ANN). This combination improves the predictive accuracy and robustness of the control system against the inherent nonlinearities and time-varying characteristics of wind energy systems. Experimental validation demonstrated that the ANN-identified MPC achieved robust voltage stabilization, limiting peak overshoot to approximately 3.5% and ensuring a settling time of 0.7 seconds under transient conditions. These results underscore the significant improvements in control precision and stability offered by the proposed approach, highlighting its potential for practical implementation in improving the reliability and power quality of wind energy conversion systems.