Deep Learning based PQD Classification using Time and Frequency Domain Features

Maintaining a continuous and reliable electric supply is vital for modern society to function effectively. In maintaining our electrical grid, power quality disturbances(PQDs) result in damage, data loss, and financial loss. Ensuring grid resilience and reducing their impact require accurate identification. PQDs, even little ones, can have serious long-term effects. Improved methods of deep learning for detection, categorization, and prevention are made possible by deep learning, signal processing, and smart grid technologies. In order to improve fault classification, this research suggests a novel method that combines temporal and frequency information using a CNN. Findings demonstrate how well it can identify PQDs, allowing for improved PQD classification in power networks.