Non-Intrusive Load Monitoring and Forecasting for Home Appliances using Artificial Intelligence – A Review

Abstract

Non-intrusive load monitoring (NILM) provides insights into how much energy consumers are consuming, encouraging them to make energy-saving changes. Load forecasting, on the other hand, provides information for accurate demand planning, grid management, and predictive maintenance of appliances. Load forecasting certainly brings more value to a NILM system. Even so, very few studies combine these two factors. Therefore, there is no review of the existing literature that addresses both NILM and load forecasting, leading to the establishment of this review. In addition, this paper also focuses on appliance-level load forecasting, implementation or simulations of non-intrusive load monitoring and forecasting, and summary and analysis of the reviewed studies. The studies were reviewed based on the motivation, dataset, data processing, feature extraction, AI/ML Techniques, implementation, and future work.