A residential load identification algorithm based on periodogram for non-intrusive load monitoring

Abstract

Non-intrusive load monitoring is an important development direction of electric load monitoring. Traditional NILM mainly track and decompose the power voltage current, and other transient or steady-state parameters like power parameters at the monitoring system entrance, and then identify the specific type of load. The load identification method is complicated, and there are limitations on identifying multiple devices switching simultaneously. This paper studied on NILM oriented residential load identification methods and proposed one based on periodogram, and then made the detailed theoretical analysis and simulation verification of the selection and determination of electric appliances characteristic model. This method could greatly simplify the system complexity and improve the efficiency of identification. It offers important reference value for the research on load online monitoring methods in residential domain.