Load identification in nonintrusive load monitoring using steady-state and turn-on transient energy algorithms

Abstract

Non-intrusive load monitoring (NILM) techniques are based on the analysis of load energy signatures. With characterizing associated transient energy signature, the reliability and accuracy of recognition results can be accurately understood or ascertained. In this study, the computer supported cooperative work techniques (CSCW), artificial neural networks (ANN), in combination with turn-on transient energy analysis, are used to identify loads and to improve recognition accuracy and computational speed of NILM results. The experimental results indicated that the incorporation of turn-on transient energy signature analysis into NILM revealed more information than traditional NILM methods, and the resulting recognition accuracy and computational speed were improved. In addition, in combination with computer supported cooperative work in electromagnetic transient program (EMTP) simulation, calculations of turn-on transient energy facilitated load identification that had significant effect on NILM results.