Skip to Main Content
Due to the global warming and climate change, it is very important to effectively improve the efficiency of the electricity energy consumption. Monitoring the power consumption of residences and buildings is one of the approaches that can improve the efficiency of the electricity energy consumption. In this paper, a Non-Intrusive Appliance Load Monitoring (NIALM) system, which applies a neuro-fuzzy pattern recognizer (NFPR) with Linguistic Hedges (LHs) to recognize the operation status of individual appliances, is proposed. A two-stage fuzzy pattern recognition process is presented in this paper. First, Fuzzy C-Means (FCM) clustering is employed to coarsely estimate the parameters used in NFPR. Following this stage, the Scaled Conjugate Gradient (SCG) training algorithm is applied to adaptively fine tune the parameters. In the proposed NIALM system, either load energizing or load de-energizing transient features are extracted from an acquired transient current waveform. NFPR performs load recognition based on these transient features. The recognition results obtained from different real experimental environments confirm that the proposed approach is able to identify the operational status of individual appliances.