Paper Key : IRJ************197
Author: Mohammad Ahsan, Sachindra Kumar Verma
Date Published: 09 Jul 2024
Abstract
Power systems often encounter faults that can result in the damage of expensive components such as motors, generators, and transformers, as well as pose risks such as over-voltages, high currents, outages, and even fatalities. To mitigate these issues, a power protection system is necessary to quickly detect, classify, and locate faults in order to minimize their impact. Analyzing power system faults is crucial for ensuring uninterrupted power supply, reducing disruptions, and preventing equipment damage. This technical description presents a detailed methodology for detecting, classifying, and determining the fault location in power systems. The primary objectives of this study are to accurately detect and classify various types of faults occurring at different locations and resistance levels, gain insights into the causes of interruptions, promptly restore power, and minimize future occurrences. Additionally, the analysis aims to improve understanding of the protection system components to implement preventive measures and decrease the likelihood of service disruptions and equipment damage. The proposed solution utilizes simple neural networks, which are effective machine learning models, for precise fault detection, classification, and fault location determination. By leveraging the capabilities of neural networks, the methodology achieves accurate results.