DATA LABELING FOR POWER SYSTEM OPERATION MONITORING

Abstract

During the process of collecting data for power system operation monitoring, the received data is often compiled in tabular form. This is the most common way to arrange data which makes easier to manipulate, analyze, and manage data. However, data tabular has characteristics such as missing or noisy data, many categories, many different values and difficult to apply Transfer learning that cause makes it difficult to monitor the operation of the power system with a huge amount of data. To monitor operations in real time, system data is often included in operational monitoring software with the new trend of using Artificial intelligence (AI) technology. In which, a complete database is structured and properly labeled, it is the most important and time-consuming step that helps the system provide accurate results, and the module identifies factors to train the model to achieve the best results. This study has built processes, labeling library set and labeled power system data for deployment purposes of power system operation monitoring models