Supervised-learning-based hour-ahead demand response for a behavior-based home energy management system approximating MILP optimization

Huy Truong Dinh, Kyu-haeng Lee and Daehee Kim

Applied Energy, 2022, vol. 321, issue C, No S0306261922007231

Abstract: The demand response (DR) program of a traditional home energy management system (HEMS) usually controls or schedules appliances to monitor energy usage, minimize energy cost, and maximize user comfort. In this study, instead of interfering with appliances and changing residents’ behavior, the proposed hour-ahead DR strategy first learns the appliance usage behavior of residents; subsequently, based on this knowledge, it silently controls the energy storage system (ESS) and renewable energy system (RES) to minimize the daily energy cost. To accomplish the goal, the proposed deep neural networks (DNNs) of this DR approximate the MILP optimization using supervised learning. The training datasets are created from the optimal outputs of an MILP solver using historical data. After training, in each time slot, these DNNs are used to control the ESS and RES using the real-time data of the surrounding environment. For comparison, we develop two different strategies, namely, the multi-agent reinforcement learning-based strategy, which is an hour-ahead strategy, and the forecast-based MILP strategy, which is a day-ahead strategy. For evaluation and verification, the proposed approaches are applied to three different real-world homes with real-world real-time global horizontal irradiation and prices. Numerical results verify the effectiveness and superiority of the proposed MILP-based supervised learning strategy, in terms of the daily energy cost.

Keywords: Behavior-based HEMS; MILP; Supervised learning; Deep reinforcement learning; Demand response (search for similar items in EconPapers)
Date: 2022
References: View references in EconPapers View complete reference list from CitEc
Citations: View citations in EconPapers (9)

Downloads: (external link)
http://www.sciencedirect.com/science/article/pii/S0306261922007231
Full text for ScienceDirect subscribers only

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:eee:appene:v:321:y:2022:i:c:s0306261922007231

Ordering information: This journal article can be ordered from
http://www.elsevier.com/wps/find/journaldescription.cws_home/405891/bibliographic
http://www.elsevier. ... 405891/bibliographic

DOI: 10.1016/j.apenergy.2022.119382

Access Statistics for this article

Applied Energy is currently edited by J. Yan

More articles in Applied Energy from Elsevier
Bibliographic data for series maintained by Catherine Liu ().