Power-Spectrum-Based Wavelet Transform for Nonintrusive Demand Monitoring and Load Identification

Abstract

Though the wavelet transform coefficients (WTCs) contain plenty of information needed for turn-on/off transient signal identification of load events, adopting the WTCs directly requires longer computation time and larger memory requirements for the nonintrusive load monitoring identification process. To effectively reduce the number of WTCs representing load turn-on/off transient signals without degrading performance, a power spectrum of the WTCs in different scales calculated by Parseval's theorem is proposed and presented in this paper. The back-propagation classification system is then used for artificial neural network construction and load identification. The high success rates of load event recognition from both experiments and simulations have proved that the proposed algorithm is applicable in multiple load operations of nonintrusive demand monitoring applications.