Automatic Fault Detection in Industrial Smart Grids Using KNN and Ensemble Classifiers

The use of sensitive electrical gadgets in industries, buildings, smart cities, and homes has increased drastically recent years. PQ events such as interruptions, surges, sags have a high impact on these devices. failure delicate devices real-time applications, particularly may result significant damage. supply quality decreases because the internal transmission system elements, unbalanced loads, other outdoor issues like weather. Several academics proposed techniques to analyze disturbances, including wavelet packets, S-transform, rough sets neural networks. In all available algorithms, classification procedure involves extraction large set features from transformed outputs, training classifier, finally making conclusion with classifier. Because involvement number features, computational cost methods increases. To reduce complexity enhance efficiency, method focuses extracting fewer low-complexity signals. Pattern recognition (PR) methods, wide variety K-nearest neighbors (KNN) ensemble classifiers, are used classify this study. performance ML approaches' is evaluated at various testing rates. Subsequently, strategies was compared that current determine dominance approaches.