Abstract
Non-Intrusive Load Monitoring (NILM) increases awareness on user energy usage patterns. In this paper, an efficient and highly accurate NILM method is proposed featuring condensed representation, super-state and fusion of two deep learning based models. Condensed representation helps the two models perform more efficiently and preserve longer-term information, while super-state helps the model to learn correlations between appliances. The first model is a deep user model that learns user appliances usage patterns to predict the next appliance usage behavior based on past behaviors by capturing the dynamics of user behaviors history and appliances usage habits. The second model is a deep appliance group model that learns the characteristics of appliances with temporal and electrical information. These two models are then fused to perform NILM. The case study based on REFIT datasets demonstrates that the proposed NILM method outperforms two state-of-the-art benchmark methods.
Highlights
Buildings consume more than 76% electricity in the United States, which can be reduced up to 15–40% using a home energy management system (HEMS) [1]
The Non-Intrusive Load Monitoring (NILM) problem is reformulated into a sequence-to-symbols problem, in which the super-state in each time interval is the symbol to be inferred from sequence of condensed representations
The training process is summarized as follows: During inference, first, the bins in (1) are inferred by the appliance group model according to power consumption and temporal information, and the user model adjusts the output of the appliance group model according to the dynamics of user behaviors history and appliances usage habits
Summary
Buildings consume more than 76% electricity in the United States, which can be reduced up to 15–40% using a home energy management system (HEMS) [1]. In order to realize a more effective HEMS that considers both users’ comfort preference and appliances’. Operation flexibility [2], the HEMS must be provided with detailed energy consumption information. Non-Intrusive Load Monitoring (NILM) is a cost-effective approach to provide such detailed energy consumption of individual appliances based on only the aggregated power measured by a single upstream power meter. NILM requires appliance-level electricity consumption to learn a machine learning model in learning phase, and the down-stream sensors are removed during operation [3,4,5,6,7,8,9,10]
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have