Recognition of Multiple Power Quality Disturbances Using KNN-Bayesian

Abstract

This paper presents a novel approach using multi-label classification to predict the classes of multiple disturbances for a power quality (PQ) event in power systems. A novel multi-label lazy learning approach named KNN-Bayesian is designed to recognize the multiple power quality disturbances based on k-nearest neighbor and Bayesian methods. In terms of the especial multi-label evaluation metrics, the proposed approach is tested on powerline signals that contain single and multiple disturbances. Experimental results indicate that KNN-Bayesian can effectively classify multiple PQ disturbances under different noise conditions