Parameter identifiability in Bayesian inference for building energy models

被引:18
|
作者
Yi, Dong Hyuk [1 ]
Kim, Deuk Woo [2 ]
Park, Cheol Soo [3 ]
机构
[1] Seoul Natl Univ, Coll Engn, Dept Architecture & Architectural Engn, 1 Gwanak Ro, Seoul 08826, South Korea
[2] Korea Inst Civil Engn & Bldg Technol, Dept Living & Built Environm Res, 283 Goyang Daero, Goyang 10223, Gyeonggi, South Korea
[3] Seoul Natl Univ, Coll Engn, Dept Architecture & Architectural Engn, Inst Construct & Environm Engn,Inst Engn Res, 1 Gwanak Ro, Seoul 08826, South Korea
关键词
Bayesian inference; Parameter identifiability; Likelihood confidence interval; Likelihood confidence region; Biplot; UNCERTAINTY ANALYSIS; RETROFIT ANALYSIS; CALIBRATION; PERFORMANCE; STOCK; DISTRIBUTIONS; CONSUMPTION; SIMULATION;
D O I
10.1016/j.enbuild.2019.06.012
中图分类号
TU [建筑科学];
学科分类号
0813 ;
摘要
Parameter identifiability is the concept of whether uncertain parameters can be correctly estimated from the observed data. The main cause of parameter unidentifiability in Bayesian inference is known as 'over-parameterization'. In this study, the likelihood confidence interval (CI) and the likelihood confidence region (CR) were introduced for quantifying the parameter identifiability. The likelihood Cl and CR can be regarded as the parameter range (one-dimensional) and parameter space (two-dimensional or higher) that can identify parameter values, respectively. For this purpose, an EnergyPlus reference office building provided by the US DOE was used in this study. Four estimation parameters in the EnergyPlus model were analyzed using the likelihood CI and CR. It was found that the closer the likelihood CI of a parameter is to the prior's parameter range, the more unidentifiable the parameter. In addition, a biplot analysis was conducted to examine a correlation between two parameters. The more correlated a parameter is with others, the more unidentifiable the parameter. It is suggested that the visual assessment of likelihood Cls and CRs can help in investigating whether Bayesian inference results can be accurately obtained. (C) 2019 Elsevier B.V. All rights reserved.
引用
收藏
页码:318 / 328
页数:11
相关论文
共 50 条
  • [1] Model selection for parameter identifiability problem in Bayesian inference of building energy model
    Yi, Dong Hyuk
    Park, Cheol Soo
    [J]. ENERGY AND BUILDINGS, 2021, 245
  • [2] Practical parameter identifiability and handling of censored data with Bayesian inference in mathematical tumour models
    Porthiyas, Jamie
    Nussey, Daniel
    Beauchemin, Catherine A. A.
    Warren, Donald C.
    Quirouette, Christian
    Wilkie, Kathleen P.
    [J]. NPJ SYSTEMS BIOLOGY AND APPLICATIONS, 2024, 10 (01)
  • [3] BAYESIAN INFERENCE CALIBRATION OF BUILDING ENERGY MODELS FOR ARID WEATHER
    Hou, Danlin
    Hassan, Ibrahim Galal
    Wang, Liangzhu
    [J]. PROCEEDINGS OF THE 2021 ASME VERIFICATION AND VALIDATION SYMPOSIUM (VVS2021), 2021,
  • [4] Parameter Identifiability for Multi-Zone Building Models
    Agbi, Clarence
    Song, Zhen
    Krogh, Bruce
    [J]. 2012 IEEE 51ST ANNUAL CONFERENCE ON DECISION AND CONTROL (CDC), 2012, : 6951 - 6956
  • [5] Determination of parameter identifiability in nonlinear biophysical models: A Bayesian approach
    Hines, Keegan E.
    Middendorf, Thomas R.
    Aldrich, Richard W.
    [J]. JOURNAL OF GENERAL PHYSIOLOGY, 2014, 143 (03): : 401 - 416
  • [6] Inference-based assessment of parameter identifiability in nonlinear biological models
    Daly, Aidan C.
    Gavaghan, David
    Cooper, Jonathan
    Tavener, Simon
    [J]. JOURNAL OF THE ROYAL SOCIETY INTERFACE, 2018, 15 (144)
  • [7] Two case studies detailing Bayesian parameter inference for dynamic energy budget models
    Boersch-Supan, Philipp H.
    Johnson, Leah R.
    [J]. JOURNAL OF SEA RESEARCH, 2019, 143 : 57 - 69
  • [8] Parameter optimisation using Bayesian inference for spallation models
    Hirtz, J.
    David, J. -c.
    Cugnon, J.
    Leya, I.
    Rodriguez-Sanchez, J. L.
    Schnabel, G.
    [J]. EUROPEAN PHYSICAL JOURNAL A, 2024, 60 (07):
  • [9] Bayesian parameter inference for models of the Black and Scholes type
    Gzyl, Henryk
    ter Horst, Enrique
    Malone, Samuel W.
    [J]. APPLIED STOCHASTIC MODELS IN BUSINESS AND INDUSTRY, 2008, 24 (06) : 507 - 524
  • [10] PARAMETER IDENTIFIABILITY IN SPARSE BAYESIAN LEARNING
    Pal, Piya
    Vaidyanathan, P. P.
    [J]. 2014 IEEE INTERNATIONAL CONFERENCE ON ACOUSTICS, SPEECH AND SIGNAL PROCESSING (ICASSP), 2014,