Please use this identifier to cite or link to this item:
|Title:||DEVELOPMENT OF FRAMEWORK BASED ALGORITHMS FOR NON-INTRUSIVE DISAGGREGATION OF PLUG LOADS FOR OFFICES||Authors:||KALLURI MALLIKARJUNA BALAJI||Keywords:||office buildings, non-intrusive load monitoring, office plug load dataset, subsequence mining, supervised classification||Issue Date:||13-Jan-2017||Citation:||KALLURI MALLIKARJUNA BALAJI (2017-01-13). DEVELOPMENT OF FRAMEWORK BASED ALGORITHMS FOR NON-INTRUSIVE DISAGGREGATION OF PLUG LOADS FOR OFFICES. ScholarBank@NUS Repository.||Abstract:||Plug load audit in a typical office environment is not only challenging, but also expensive and time-consuming. The proposed data-driven framework aims at improving existing state-of-the-art protocol to monitor plug loads in every individual workstation. It measures the aggregated energy signature of multiple office appliances (e.g. desktop PC, monitor, laptop PC and multi-functional device) and employs intelligent algorithms to disambiguate operational modes of every individual ground-truth appliance. It reduces both capital cost in metering every appliance and also offers valuable insights from single-point measurement. The analysis is driven based on a large repository called OPLD developed for this purpose. A novel idea of transforming temporal energy signatures into a massive dictionary of Context-Free-Grammar words is implemented within this framework. Such implementation facilitates developing multiple load disaggregation algorithms based on substring similarity. Several potential aggregate appliance use-cases are analyzed. The results of deep disaggregation are benchmarked and validated across multiple algorithms.||URI:||http://scholarbank.nus.edu.sg/handle/10635/135724|
|Appears in Collections:||Ph.D Theses (Open)|
Show full item record
Files in This Item:
|KalluriMB.pdf||6.54 MB||Adobe PDF|
checked on Jun 6, 2019
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.