Online HVAC system modeling with BMS data using recurrent neural networks

E. Togashi; S. Tanabe

Online HVAC system modeling with BMS data using recurrent neural networks

E. Togashi^*, S. Tanabe

^*Corresponding author for this work

School of Creative Science and Engineering

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

1 Citation (Scopus)

Abstract

This paper presents a method to develop and tune HVAC system models with BMS data using recurrent neural networks (RNN). To update RNN models automatically online, we eliminate ad hoc adjustment technique in three points. One is in selection of inputs variables. We propose a method of calculating the necessity of input variables with using connections' weights of neurons. Second is in the selection of hidden layers unit numbers. We introduce "Cross-validation" method and it gives criterion to evaluate parametric model. Last is in selection of training data. We apply cluster analysis to BMS data to evaluate scarcity of data. Proposed approaches are tested real measurements of actual buildings in Japan.

Original language	English
Title of host publication	HB 2006 - Healthy Buildings
Subtitle of host publication	Creating a Healthy Indoor Environment for People, Proceedings
Pages	407-410
Number of pages	4
Publication status	Published - 2006 Dec 1
Event	Healthy Buildings: Creating a Healthy Indoor Environment for People, HB 2006 - Lisboa, Portugal Duration: 2006 Jun 4 → 2006 Jun 8

Publication series

Name	HB 2006 - Healthy Buildings: Creating a Healthy Indoor Environment for People, Proceedings
Volume	4

Conference

Conference	Healthy Buildings: Creating a Healthy Indoor Environment for People, HB 2006
Country/Territory	Portugal
City	Lisboa
Period	06/6/4 → 06/6/8

Keywords

BMS
Online modeling
Recurrent neural networks

ASJC Scopus subject areas

Civil and Structural Engineering
Building and Construction
Health, Toxicology and Mutagenesis

Cite this

Online HVAC system modeling with BMS data using recurrent neural networks. / Togashi, E.; Tanabe, S.
HB 2006 - Healthy Buildings: Creating a Healthy Indoor Environment for People, Proceedings. 2006. p. 407-410 (HB 2006 - Healthy Buildings: Creating a Healthy Indoor Environment for People, Proceedings; Vol. 4).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

Togashi, E & Tanabe, S 2006, Online HVAC system modeling with BMS data using recurrent neural networks. in HB 2006 - Healthy Buildings: Creating a Healthy Indoor Environment for People, Proceedings. HB 2006 - Healthy Buildings: Creating a Healthy Indoor Environment for People, Proceedings, vol. 4, pp. 407-410, Healthy Buildings: Creating a Healthy Indoor Environment for People, HB 2006, Lisboa, Portugal, 06/6/4.

@inproceedings{0b7e8d4612cc4d55beeb32ee03294cf2,

title = "Online HVAC system modeling with BMS data using recurrent neural networks",

abstract = "This paper presents a method to develop and tune HVAC system models with BMS data using recurrent neural networks (RNN). To update RNN models automatically online, we eliminate ad hoc adjustment technique in three points. One is in selection of inputs variables. We propose a method of calculating the necessity of input variables with using connections' weights of neurons. Second is in the selection of hidden layers unit numbers. We introduce {"}Cross-validation{"} method and it gives criterion to evaluate parametric model. Last is in selection of training data. We apply cluster analysis to BMS data to evaluate scarcity of data. Proposed approaches are tested real measurements of actual buildings in Japan.",

keywords = "BMS, Online modeling, Recurrent neural networks",

author = "E. Togashi and S. Tanabe",

year = "2006",

month = dec,

day = "1",

language = "English",

isbn = "9789899506718",

series = "HB 2006 - Healthy Buildings: Creating a Healthy Indoor Environment for People, Proceedings",

pages = "407--410",

booktitle = "HB 2006 - Healthy Buildings",

note = "Healthy Buildings: Creating a Healthy Indoor Environment for People, HB 2006 ; Conference date: 04-06-2006 Through 08-06-2006",

}

TY - GEN

T1 - Online HVAC system modeling with BMS data using recurrent neural networks

AU - Togashi, E.

AU - Tanabe, S.

PY - 2006/12/1

Y1 - 2006/12/1

N2 - This paper presents a method to develop and tune HVAC system models with BMS data using recurrent neural networks (RNN). To update RNN models automatically online, we eliminate ad hoc adjustment technique in three points. One is in selection of inputs variables. We propose a method of calculating the necessity of input variables with using connections' weights of neurons. Second is in the selection of hidden layers unit numbers. We introduce "Cross-validation" method and it gives criterion to evaluate parametric model. Last is in selection of training data. We apply cluster analysis to BMS data to evaluate scarcity of data. Proposed approaches are tested real measurements of actual buildings in Japan.

AB - This paper presents a method to develop and tune HVAC system models with BMS data using recurrent neural networks (RNN). To update RNN models automatically online, we eliminate ad hoc adjustment technique in three points. One is in selection of inputs variables. We propose a method of calculating the necessity of input variables with using connections' weights of neurons. Second is in the selection of hidden layers unit numbers. We introduce "Cross-validation" method and it gives criterion to evaluate parametric model. Last is in selection of training data. We apply cluster analysis to BMS data to evaluate scarcity of data. Proposed approaches are tested real measurements of actual buildings in Japan.

KW - BMS

KW - Online modeling

KW - Recurrent neural networks

UR - http://www.scopus.com/inward/record.url?scp=84871587091&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84871587091&partnerID=8YFLogxK

M3 - Conference contribution

AN - SCOPUS:84871587091

SN - 9789899506718

T3 - HB 2006 - Healthy Buildings: Creating a Healthy Indoor Environment for People, Proceedings

SP - 407

EP - 410

BT - HB 2006 - Healthy Buildings

T2 - Healthy Buildings: Creating a Healthy Indoor Environment for People, HB 2006

Y2 - 4 June 2006 through 8 June 2006

ER -

Online HVAC system modeling with BMS data using recurrent neural networks

Abstract

Publication series

Conference

Keywords

ASJC Scopus subject areas

Other files and links

Fingerprint

Cite this