Accurate quantitative estimation of energy performance of residential buildings using statistical machine learning tools

被引:532
作者
Tsanas, Athanasios [1 ]
Xifara, Angeliki [2 ]
机构
[1] Univ Oxford, Inst Math, OCIAM, Oxford OX1 3LB, England
[2] Cardiff Univ, Architectural Sci Grp, Welsh Sch Architecture, Cardiff, S Glam, Wales
基金
英国工程与自然科学研究理事会;
关键词
Building energy evaluation; Heating load; Cooling load; Non-parametric statistics; Statistical machine learning; CONSUMPTION; SIMULATION; STANDARD;
D O I
10.1016/j.enbuild.2012.03.003
中图分类号
TU [建筑科学];
学科分类号
0813 ;
摘要
We develop a statistical machine learning framework to study the effect of eight input variables (relative compactness, surface area, wall area, roof area, overall height, orientation, glazing area, glazing area distribution) on two output variables, namely heating load (HL) and cooling load (CL), of residential buildings. We systematically investigate the association strength of each input variable with each of the output variables using a variety of classical and non-parametric statistical analysis tools, in order to identify the most strongly related input variables. Then, we compare a classical linear regression approach against a powerful state of the art nonlinear non-parametric method, random forests, to estimate HL and CL Extensive simulations on 768 diverse residential buildings show that we can predict HL and CL with low mean absolute error deviations from the ground truth which is established using Ecotect (0.51 and 1.42, respectively). The results of this study support the feasibility of using machine learning tools to estimate building parameters as a convenient and accurate approach, as long as the requested query bears resemblance to the data actually used to train the mathematical model in the first place. (C) 2012 Elsevier B.V. All rights reserved.
引用
收藏
页码:560 / 567
页数:8
相关论文
共 26 条
  • [1] [Anonymous], 2003, OFFICIAL J EURO 0401, pL1/65
  • [2] Bishop C. M., 2007, Technometrics, DOI DOI 10.1198/TECH.2007.S518
  • [3] Random forests
    Breiman, L
    [J]. MACHINE LEARNING, 2001, 45 (01) : 5 - 32
  • [4] China building energy consumption: Situation, challenges and corresponding measures
    Cai, W. G.
    Wu, Y.
    Zhong, Y.
    Ren, H.
    [J]. ENERGY POLICY, 2009, 37 (06) : 2054 - 2059
  • [5] Development and validation of regression models to predict monthly heating demand for residential buildings
    Catalina, Tiberiu
    Virgone, Joseph
    Blanco, Eric
    [J]. ENERGY AND BUILDINGS, 2008, 40 (10) : 1825 - 1832
  • [6] Chatfield C., 2004, The analysis of time series: An introduction
  • [7] Cover T. M., 1999, Elements of information theory
  • [8] Contrasting the capabilities of building energy performance simulation programs
    Crawley, Drury B.
    Hand, Jon W.
    Kurnmert, Michal
    Griffith, Brent T.
    [J]. BUILDING AND ENVIRONMENT, 2008, 43 (04) : 661 - 673
  • [9] Applying support vector machines to predict building energy consumption in tropical region
    Dong, B
    Cao, C
    Lee, SE
    [J]. ENERGY AND BUILDINGS, 2005, 37 (05) : 545 - 553
  • [10] Hastie T., 2009, ELEMENTS STAT LEARNI, DOI DOI 10.1007/978-0-387-84858-7