Non-Intrusive Load Monitoring using Electricity Smart Meter Data: A Deep Learning Approach

Abstract

Non-Intrusive Load Monitoring (NILM) is the process of decomposing an aggregated building electricity mains measurement into individual appliances. NILM is a very challenging classification problem and a number of statistical techniques have been proposed for this. Recent advances have made deep learning a dominant approach for classification in fields such as image processing and speech recognition. This paper investigates the application of deep learning approaches in NILM, and develops a NILM classifier that can detect the activations of common electrical appliances from smart meter data. The performance of the NILM deep learning classifier is demonstrated using publicly- available smart meter data sets, and the ability of the classifier to generalise to unseen data is examined.