Electrical Power Quality Disturbances Detection in Transmission Lines Using Machine Learning-Enabled Classifier

There are increasing concerns about power quality disturbances (PQDs) at many phases of energy generation, transformation, distribution, and consumption due to the increasing interconnection of various energy systems. The basis for addressing PQDs is the automatic categorization of voltage or phase angle disturbances. According to a usual standpoint, the three distinct steps of signal analysis, feature selection, and classification should be used to separate the detection of issues with power quality. Nevertheless, signal analysis possesses several inherent deficiencies, mostly stemming from the laborious and inaccurate process of human feature selection. Consequently, this results in diminished classification accuracy when dealing with many disturbances and a compromised ability to withstand disruptive interference. This study focuses on the identification and categorization of PQD using a machine learning-based classifier, taking into account the features of the power quality problems problem, eight different features are taken from the voltage data and used to figure out what caused each PQD. Various types of machine learning models are employed to analyze the dataset, and the effectiveness of the machine learning classifier is assessed by validating its performance using a separate test dataset. Once the machine learning classifier model can classify the disturbances types with 96% accuracy. The proposed classifiers can effectively detect disturbances in the transmission line.