A statistical approach to modeling and forecast of the building energy consumption

A. Vaghefi, M. A. Jafari, Jianmin Zhu, J. Brouwer, Y. Lu

Research output: Contribution to conferencePaperpeer-review

Abstract

The objective of this study is to investigate a statistical approach to model the major sources of variation in building energy consumption dynamics based upon a set of control and environmental (input) variables. The proposed statistical model is capable of forecasting energy consumption for different set points during a given time window. In this paper, two linear models are proposed to determine the relationship between energy consumption and input variables. The first model is a hybrid time-series regression model that is a combination of a multiple linear regression model and a seasonal autoregressive moving average (ARMA) model. The second model is a Box-Jenkins transfer function model, which is often used to explain the relation between variables of a dynamic linear time-invariant system. The proposed models can either be treated as an approximation to EnergyPlus, energy simulation software, or as a model for the optimization of set points for buildings according to specifications of building operators. The performance of the proposed methodology is evaluated through a use-case analysis on an existing building.

Original languageEnglish (US)
Pages2838-2845
Number of pages8
StatePublished - 2013
EventIIE Annual Conference and Expo 2013 - San Juan, Puerto Rico
Duration: May 18 2013May 22 2013

Other

OtherIIE Annual Conference and Expo 2013
Country/TerritoryPuerto Rico
CitySan Juan
Period5/18/135/22/13

All Science Journal Classification (ASJC) codes

  • Industrial and Manufacturing Engineering

Keywords

  • ARMA model
  • Box-Jenkins model
  • EnergyPlus
  • Time series regression model

Fingerprint

Dive into the research topics of 'A statistical approach to modeling and forecast of the building energy consumption'. Together they form a unique fingerprint.

Cite this