Windowing Based Continuous S-Transform (ST) with Deep Learning for Detection and Classifying Power Quality Disturbances (PQDs)

Abstract

Transform (ST) with deep learning classifier for interrupt and transient disturbance detection and classification.  Interruption is a type of disturbance in power quality (PQ).  The main goal is  to analyze the detection and classification of voltage interrupt and transient using ST as a signal processing technique. Windowing techniques are classified into half-cycle and one-cycle windowing techniques (WT), with both cycles being utilized for comparison. The MATLAB programming language was used to construct the disturbance signal, which was saved as an m-file. The significant feature in the form of scattering data was extracted from the disturbance signal using ST. The scattering data was then used to create a detection interface within the disturbance signal. The scattering data is fed into a neural network (NN) that classifies the disturbance signal's percentage accuracy. This study provides a windowing technique which can provide smooth detection and adequate characteristics for high accuracy percentages in power quality disturbance classification (PQDs).