A Non-Intrusive Load Monitoring (NILM) method for residential appliances based on ac-\udtive power signal is presented. This method works e\udectively with a single active power measurement\udtaken at a low sampling rate (1 s). The proposed method utilizes the\udKarhunen Lo\ud ́\udeve\ud(KL) expan-\udsion to decompose windows of active power signals into subspace components in order to construct a\udunique set of features, referred to as signatures, from individual and aggregated active power signals.\udSimilar signal windows were clustered in to one group prior to feature extraction. The clustering was\udperformed using a modified mean shift algorithm. After the feature extraction, energy levels of signal\udwindows and power levels of subspace components were utilized to reduce the number of possible ap-\udpliance combinations and their energy level combinations. Then, the turned on appliance combination\udand the energy contribution from individual appliances were determined through the Maximum a Pos-\udteriori (MAP) estimation. Finally, the proposed method was modified to adaptively accommodate the\udusage patterns of appliances at each residence. The proposed NILM method was validated using data\udfrom two public databases:\udtracebase\udand reference energy disaggregation data set (REDD). The pre-\udsented results demonstrate the ability of the proposed method to accurately identify and disaggregate\udindividual energy contributions of turned on appliance combinations in real households. Furthermore,\udthe results emphasise the importance of clustering and the integration of the usage behaviour pattern in\udthe proposed NILM method for real households
展开▼