Identification of Multi-Sinusoidal Signal Models: Application to Non-Intrusive Industrial Load Monitoring

This paper proposes a new Non-Intrusive Load Monitoring (NILM) approach to deal with industrial settings. The goal is to supervise the states of a set of appliances connected to the same electrical network. To do so, the proposed method makes use of characteristic frequencies, thus implying the need of an algorithm capable of estimating signal frequencies and magnitudes. Moreover, prior information about the appliances is to be obtained through a feature extraction step. Some classification algorithms will be tested and compared to each other.