Disaggregated End-Use Energy Sensing for the Smart Grid

TLDR

Most utility meters only provide aggregate consumption, whereas disaggregated data that identifies individual device usage offers a richer dataset for homeowners, operators, utilities, and policymakers. This article surveys current and emerging disaggregation methods and identifies signal features that could enable cost‑effective, viable disaggregated sensing. The authors present a voltage‑noise based load‑disaggregation approach, detailing its sensing hardware, classification algorithms, and evaluation across 14 homes. The paper concludes by outlining open research challenges that must be overcome for widespread deployment of disaggregated energy sensing.

Abstract

Most energy meters installed by utilities are intended primarily to support billing functions. Meters report only the aggregate energy consumption of a home or business over intervals as long as a month. In contrast, disaggregated energy usage data identified by individual devices or appliances offers a much more descriptive dataset that has the potential to inform and empower a wide variety of energy stakeholders, from homeowners and building operators to utilities and policy makers. In this article, the authors survey existing and emerging disaggregation techniques and highlight signal features that might be used to sense disaggregated data in a viable and cost-effective manner. They provide a summary of a new approach to electrical load disaggregation that uses voltage noise, including a brief overview of their sensing hardware, classification algorithms, and evaluation in 14 homes. The article concludes with a discussion of current open research problems that must be addressed before disaggregated energy sensing can be widely deployed.