Abnormal Voltage Regulation Detection in on-Grid PV-ESS System by Support Vector Machine with Principal Component Analysis

This paper describes an abnormal voltage regulation detection approach based on machine learning algorithm in a on-Grid photovoltaic (PV) system. The three-phase AC voltage of the power conversion system is measured using a voltage transformer and used as a feature of the proposed detection scheme. We integrated the principle component analysis (PCA) technique and the support vector machine to improve the algorithm’s accuracy and reliability (SVM). Before applying the real-time data to the SVM model, the PCA analysis is performed. The proposed system’s implementation results demonstrate its efficacy and robustness in the intelligent PV-ESS energy farm.