Learning Individual Preferences for Energy-Efficiency and Comfortable Living

Chayan Sarkar, SN Akshay Uttama Nambi, R. Venkatesha Prasad

Research output: Chapter in Book/Conference proceedings/Edited volumeConference contributionScientificpeer-review

Abstract

Automatic control of HVAC and artificial lights has been one of the popular methods for achieving energy-efficient buildings. The operating set-points are decided based on predefined values to ensure comfort level to most of the occupants based on prior studies. However, a person can feel comfortable beyond the traditional set-point ranges used in the energy management systems of buildings. In this work, we develop a smartphone application that learns individual preferences about thermal and visual comfort with minimal user intervention. These functions provide the flexibility to operate the controllers in an aggressively lower energy consuming state while maintaining the comfort level of the occupants. Using a HVAC energy consumption model, we show that individual comfort preference based set-point can attain lesser energy consumption as compared to fixed set-point.
Original languageEnglish
Title of host publicationProceedings of the 2Nd ACM International Conference on Embedded Systems for Energy-Efficient Built Environments
Place of PublicationNew York, NY, USA
PublisherACM Sigmobile
Pages115-116
Number of pages2
ISBN (Print)978-1-4503-3981-0
DOIs
Publication statusPublished - 2015
Event2Nd ACM International Conference on Embedded Systems for Energy-Efficient Built Environments - Seoul, Korea, Democratic People's Republic of
Duration: 4 Nov 20155 Nov 2015

Publication series

NameBuildSys '15
PublisherACM

Conference

Conference2Nd ACM International Conference on Embedded Systems for Energy-Efficient Built Environments
Country/TerritoryKorea, Democratic People's Republic of
CitySeoul
Period4/11/155/11/15

Fingerprint

Dive into the research topics of 'Learning Individual Preferences for Energy-Efficiency and Comfortable Living'. Together they form a unique fingerprint.

Cite this