Classification and Potential Evaluation of Residential Users in Demand Response Based on NILM Data

Power system demand response (DR) has been carried out around the world, and a large number of large-capacity users represented by industrial and commercial users have participated in it. However, the reserve resources required by the system are increasing year by year, and more resources are needed to participate in DR. Facts have proved that resident users also have the potential to respond to demand, and can participate in DR. Especially with the promotion of non-intrusive load monitoring instruments, the user data that can be obtained is more detailed. These data provide a more accurate reference for their participation in DR. In this paper, based on the non-invasive load monitoring data, the user classification technology based on Fuzzy C-Means Clustering is studied and the users with large adjustable quantity are screened out. Then three residential devices with high response potential are modeled, and their approximate response potential is given. Finally, based on the data of a residential area in Hangzhou City, Zhejiang Province, China, an example is analyzed.