Detection and classification of power quality disturbances

Abstract

Smart grid equipment has revolutionized various aspects of everyday life. However, a drawback of these systems is their nonlinear behavior, which leads to the injection of electrical disturbances into the power grid. These disturbances can result in voltage and current distortions. Artificial Intelligence (AI) techniques based on automatic feature extraction are utilized to detect and classify power quality disturbances. Subsets of AI; Deep Learning and Machine Learning algorithms such as the hybrid CNN-SVM and the Long Short-Term Memory (LSTM) are used for the detection and classification of these disturbances. The voltage signals obtained from MATLAB simulation serve as training and testing data for identification algorithms aimed at detecting various types of disturbances, such as voltage sag, voltage swell, harmonics, and interruption. The outcomes of the simulations demonstrate that the CNN-SVM technique exhibits higher accuracy in detecting and classifying power quality disturbances compared to the LSTM technique.