CLASSIFICATION OF POWER QUALITYDISTURBANCES USING WAVELET TRANSFORM ANDS-TRANSFORM BASED ARTIFICIAL NEURALNETWORK

This paper presents features that characterize power quality disturbances from recorded voltage and current signals using wavelet transformation and S-transform analysis. The disturbance of interest includes sag, swell, transient and harmonics. A 25kv distribution network has been simulated using matlab software. The feature extraction has been done using wavelet transformation and S-transform, the coefficients are collected and given to the neural network for the best classification. The S-transform based classification shows better performance in detecting, localizing and classifying compared to the wavelet transform based Back Propagation Algorithm.