Intelligent power quality monitoring by using S-transform and neural network

In this paper a method in intelligent monitoring of the power quality events is presented. The main objectives are the identification and classification of these events. A method for classification is used based on the combination of S-transform and neural networks. The S-transform, which is based on the wavelet transform with a phase correction, provides frequency dependent resolutions that simultaneously localize the real and imaginary spectra. Neural network configurations are trained with features from the S-transform for recognizing the waveform class. The whole method is tested over a variety of power network disturbance signals and their combinations which are created by EMTP simulations in a 34 bus IEEE standard network. The classification accuracy for these events is given and shows that proposed method is doing well in detecting and classifying these types of disturbances.