Towards Applicability: A Comparative Study onNon-Intrusive Load Monitoring Algorithms

Non-intrusive load monitoring is an important energy disaggregation technology, which can provide appliance-level consumption estimation given a series of total consumption over time, aiding users green awareness and living as well as device fault detection. Despite of the claimed performance, an insightful analysis on practical perspectives is still lacking. In this paper we therefore propose a unified pipeline to conduct comparative experiments on four representative state-of-the-art algorithms. We further provide an in-depth look at two crucial factors affecting algorithms’ feasibility in practice, namely sampling rate and transferability, as well as the algorithms’ performance on energy disaggregation.