Development of Power Quality Disturbances Dataset for Classification Using Deep Learning

Power quality (PQ) disturbanceDisturbances identificationIdentification is an essential and critical task for both utility and industry. Deep learning based approaches are using for automatic PQ disturbanceDisturbances that requires huge amount of data. In this chapter, the detailed procedure to develop the PQ disturbanceDisturbances dataset is discussed. Total 13 PQ disturbancesDisturbances considered while developing the dataset. Discrete Wavelet Transform (DWT) with daubechies wavelet is used to extract the featuresFeatures from PQ disturbance signal. MATLABMATLAB software is used to generate the PQ disturbance signals and to extract the detailed and approximated DWT coefficients. In this chapter, Multi-Resolution AnalysisMulti-Resolution Analysis (MRA) algorithm is used to decomposition and reconstruction the signal at resolution level of 8.