期刊
ENERGY AND BUILDINGS
卷 172, 期 -, 页码 317-327出版社
ELSEVIER SCIENCE SA
DOI: 10.1016/j.enbuild.2018.05.007
关键词
Non-intrusive approach; Data mining; Power change; Occupant energy-use behavior; Commercial buildings; Wi-Fi network; Aggregate load data
资金
- Research Council Interdisciplinary Grant Award at UNL
- Korea Agency for Infrastructure Technology Advancement [17CTAP-C128499-01]
- Korea Agency for Infrastructure Technology Advancement (KAIA) [128499] Funding Source: Korea Institute of Science & Technology Information (KISTI), National Science & Technology Information Service (NTIS)
Occupants' energy-consuming behaviors have a significant influence on overall energy consumption in commercial buildings. Accordingly, understanding and intervening in these behaviors offers a significant opportunity for energy savings in commercial buildings. Current approaches to behavior modification rely on available occupant-specific energy consumption data, but capturing such data is generally expensive. One possible solution to this challenge is to link energy consumption to individual occupants' energy-use behaviors in commercial buildings. In this context, this study proposes a non-intrusive occupant load monitoring (NIOLM) approach that couples occupancy-sensing data-captured from existing Wi-Fi infrastructureswith power changes in aggregate building-wide energy data to thereby disaggregate building-wide data down to the individual. This paper describes two case studies that investigate the feasibility of using the NIOLM approach to identify occupant-specific energy consumption information. Tracking eleven occupants' energy-use behaviors using NIOLM over a four-month period resulted in an average F-measure of 0.778 and Accuracy of 0.955. The case studies thereby demonstrated that NIOLM successfully tracks individual occupants' energy-consuming behaviors at minimal cost by utilizing existing high-resolution metering devices and Wi-Fi network infrastructures in commercial buildings. (C) 2018 Elsevier B.V. All rights reserved.
作者
我是这篇论文的作者
点击您的名字以认领此论文并将其添加到您的个人资料中。
推荐
暂无数据