Real-time system for detection and classification of power quality disturbances

Abstract

Power quality is indeed emerging as a global concern due to various factors, such as integration of new technologies, the expansion of smart grid and, increased penetration of renewables. These developments introduce new challenges that affect the stability and reliability of the electrical grid. To address the consequences of power quality issues effectively, it is crucial to detect and classify them accurately. Despite the numerous models proposed for power quality classification, achieving real−time classification remains a significant challenge. This paper presents a deep learning model for classification, evaluated using both synthetically generated data and real data. The results demonstrate excellent performance in classifying both types of data. Additionally, a real-time system for detection and classification is implemented and validated in laboratory environment. The findings indicate that the model effectively detects disturbances, thus contributing to advancements in real-time power quality.