A feasibility study of automated plug-load identification from high-frequency measurements

Abstract

Plug-meters benefit many grid and building-level energy management applications like automated load control and load scheduling. However, installing and maintaining large and/or long term deployments of such meters requires assignment and updating of the identity (labels) of electrical loads connected to them. Although the literature on electricity disaggregation and appliance identification is extensive, there is no consensus on the generalizability of the proposed solutions, especially with respect to the features that are extracted from voltage and current measurements. In this paper, we begin to address this problem by comparing the discriminative power of commonly used features. Specifically, we carry out tests on PLAID, a publicly available high-frequency dataset of hundreds of residential appliances. By examining how the classification accuracy changes with sampling frequency, we also explore the computational complexity of these techniques to understand the feasibility and design of a hardware setup that can perform these calculations in near real-time.