Detection and classification of power quality disturbances using signal processing based artificial neural networks

Abstract

Power quality refers to the unusual behavior on a power system that emerges as voltage or current and obstructs the normal operation of electrical or electronic equipment. To raise the standard of the electricity delivered, it is essential to identify the cause of the power quality issue. As a result, significant progress has been made in power quality measurement technologies. It might be difficult to identify power quality disturbance waveforms since there are so many distinct kinds of disturbances. This leads to the development of an automatic recognition system to categorize the disturbance waveforms. This thesis explores the applications of signal processing techniques and Artificial Neural Networks for the detection and classification of power quality disturbances. newlineThe Wavelet Transform (WT) decomposes the signal in the frequency domain and tracks changes in the signal over time. Besides, WT has the benefit of not requiring the signal to be stationary. Discrete Wavelet Transform (DWT) and Discrete Wavelet Packet Transform (DWPT) are used for the analysis and detection of power quality disturbances. In DWT, only low frequency bands are considered for analysis. But in DWPT, both the low and high frequency bands are considered for analysis and processing. newlineS-Transform (ST) is another important tool used for the identification and classification of power quality disturbance. It combines a resolution that is frequency dependent and simultaneously localizes the real and hypothetical spectra. Gaussian modulated co-sinusoids serve as the S-basis for finding S-Transform. Further, using S-Transform, the S-Transform contours and 3D plots are found out which are more useful for distinguishing the disturbance types. newlineThe combination of DWT and ST (DWT-ST) proposed in this work helps to analyze the signals at different frequency sub bands and produced better outcomes. Further, DWPT-ST based techniques are also proposed for the detection of disturbances. newline