Machine Learning-Driven Strategies for PMU Deployment and Fault Localization in Modern Distribution Networks

ABSTRACTThe research focused on developing machine learning-driven strategies for optimizing PMU deployment and improving faultlocalization in modern distribution networks. Phasor Measurement Units (PMUs) provide real-time data essential for monitoring power system stability and identifying faults. However, due to the high cost of PMU installation, their optimal placement in the distribution network was critical to balance coverage and cost efficiency. The study employed machine learning algorithms, including decision trees and support vector machines, to determine the most effective locations for PMUs. Additionally, various fault detection and localization methods were developed using historical data from the network. Simulation tests were conducted on a standard distribution system model, demonstrating that the machine learning approach improved fault localization accuracy compared to traditional methods. The results indicated that strategically placed PMUs significantly reduced the time and effort required to detect and isolate faults, enhancing system reliability. The findings alsohighlighted the scalability of the approach, making it suitable for both small and large distribution networks. Overall, the research contributed to more efficient monitoring of distribution systems, providing a cost-effective solution for improving fault detection and network resilience.