A Study on the Detection Method of Load Transfer in Distribution System Using Time Series Decomposition

In distribution systems, load transfers cause sudden fluctuations, necessitating rapid detection to ensure effective operations and infrastructure management. These fluctuations can compromise grid stability, and if undetected, lead to inefficiencies. Also, the growing integration of solar power has heightened the importance of identifying load transfers, especially due to reverse power flow issues. Traditional detection methods are often complex and data-intensive, limiting their practical application. This study introduces a novel method using time series decomposition, moving averages, and moving standard deviations to efficiently detect load transfers. The approach is simple, data-efficient, and offers real-time detection capabilities. Validation against real-world planned load transfers demonstrated high accuracy and reliability, proving its effectiveness in supporting grid stability and infrastructure management.