Power quality event analysis using wavelet feature based fuzzy classification

A system for power quality (PQ) disturbance (event) analysis under noisy environment is proposed using wavelet feature (WF) based fuzzy classification method. In practice, the captured signals are often corrupted by noise. Also the nonlinear and non-stationary behavior of PQ events due to various switching devices make the detection and classification tasks more cumbersome. Wavelet transform (WT) has been proven to be an effective tool for extracting inherent features through time-frequency analysis of these events. In this work we exploited WT for noise removal to make the task of detection and/or localisation of events simpler and to decompose the PQ event signals for extracting unique features. These WF are fed into fuzzy classifier for making decision regarding the type of events. We have used fuzzy product aggregation reasoning rule [8] based classifier in the present application. Varieties of PQ events which include voltage sag, swell, momentary interruption, notch, oscillatory transient and spikes are considered to test the performance of proposed approach. Comparative simulation studies revealed the superiority of the proposed method compared to WF based fuzzy explicit, fuzzy k-nearest neighbor and fuzzy maximum likelihood under noisy environment.