Cepstrum smoothing-based feature extraction method for electric loads disaggregation

This paper presents Cepstrum smoothing-based feature extraction method to reduce the variation of the characteristic signals of home appliance for Nonintrusive Load Monitoring (NILM). In this paper, a data acquisition system is also developed to obtain only characteristic signals of appliances and to filter the noise inputted from the power supply. Testing results are provided to demonstrate the effectiveness of the proposed method.