Detection of occupant actions in buildings through change point analysis of in-situ measurements

• Type: Article in International Scientific Journal
• Date: 2018

Authors

Pedro F. Pereira (Author) Other; Nuno M. M. Ramos (Author) FEUP

Journal

Title: Energy and Buildings

Vol. 173

Pages: 365-377

ISSN: 0378-7788

Publisher: Elsevier

Indexing

ISI Web of Knowledge - 5 Citations

Scopus - 18 Citations

INSPEC

Other information

Authenticus ID: P-00P-KP9

DOI: 10.1016/j.enbuild.2018.05.050

Abstract (EN): The number of research studies on occupant behaviour is increasing, following four main vectors: needs of the occupants, actions, drivers and building systems. In the specific studies which preceded this one, a large amount of indoor environment data, mainly temperature, relative humidity and carbon dioxide concentration, was collected. Moreover, the installation of specific sensors to detect occupant behaviour was surmised to not always constitute the best option, due to the complexity of its installation and maintenance, as well as its impacts on occupants. This study presents itself as a contribution to the study of residential occupant behaviour. The statistical tool of Change Point Analysis (CPA) was applied in a methodology devised to detect occupant actions in residential buildings, using data from an indoor environment monitoring system including temperature, relative humidity and carbon dioxide. The principle of this methodology is based on the fact that occupant actions produce abrupt shifts on the cited parameters. Change point analysis detects the number of abrupt changes in a time series and its precise moments, which are then correlated with occupant actions. In this work, the occupants of a dwelling in Porto, Portugal were studied. In-situ measurements were taken, as well as surveys for the assessment of the main occupant actions. The application of the proposed methodology using CPA was successful. Comparing the results of action detection with the daily journals of the occupants or specific sensors used for the same purpose, an accuracy of 97% in detecting window opening, 97% in showering, 100% in heating and 85% in cooking was obtained.

Language: English

Type (Professor's evaluation): Scientific

No. of pages: 13