Classification of Power Quality Disturbances Using Machine Learning Methods

Power quality disturbances such as voltage sags, swells, harmonics, and transients pose significant challenges to electrical systems. To address these issues, this study employs the Fourier Synchro Squeezing Transform (FSST) for preprocessing various power quality signals, enhancing the accuracy of the analysis. The research focuses on classifying these disturbances using four prominent machine learning algorithms: Support Vector Machine 91% accuracy (SVM), Random Forest 96% accuracy (RF), Gradient Boosting Machine 99% accuracy (GBM), and Logistic Regression 90% accuracy (LR). The Gradient Boosting Machine (GBM) demonstrates superior performance due to its iterative process, achieving the highest classification accuracy.