CALIBRO: an R Package for the Automatic Calibration of Building Energy Simulation Models

Monari, Filippo; Strachan, Paul

doi:10.26868/25222708.2017.224

Cited by 4 publications

(3 citation statements)

References 17 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…By applying calibration to metamodelling, the relations between building simulation and calibration, which are usually measured at the level of scene-model instances (Coakley, Raftery & Keane, 2014) are extended at design-space level. Sophisticated knowledge-based procedures have been developed for that task (Pedrini, Westphal & Lamberts, 2002;Yoon, Lee & Claridge, 2003;Raftery, Keane & O'Donnell, 2011a, 2011b, as well as methods derived from optimization (Reddy, Maor & Panjapornpon, 2007a, 2007bCoakley et al 2011) and, more recently, bayesian logic (Monari & Strachan, 2017).…”

Section: Interdisciplinary Literature Basesmentioning

confidence: 99%

Increasing the efficiency of simulation-based design explorations via metamodelling

Brunetti

2020

Journal of Building Performance Simulation

View full text Add to dashboard Cite

Building performance simulation can support parametric explorations of design option spaces.Resources available for modelling and computing often require the reduction of the descriptive information of a design solution at the level of design-space model rather than at that of building model. To obtain that reduction, two design-space models of different scope (for example, one full-scope and simplified, the other partial-scope and detailed) can be combined to make their responses usable as surrogates of those produced by a full-scope detailed inquiry. But aligning two building models of different scope can be difficult and sometimes impossible. This article presents a design-space "grafting" technique supported by metamodelling that makes it possible to hybridize two non-aligned design-space models so as to obtain a diffusely calibrated conjoint response. The strategy can be integrated into metamodelling and decomposition-based optimization to decrease the information costs entailed by parametric explorations.

show abstract

Section: Interdisciplinary Literature Basesmentioning

confidence: 99%

Increasing the efficiency of simulation-based design explorations via metamodelling

Brunetti

2020

Journal of Building Performance Simulation

View full text Add to dashboard Cite

show abstract

“…Recently, non-linear models have been proposed for predicting building energy use and other thermal quantities, e.g., Artificial Neural Networks (ANNs), Support Vector Machines (SVMs), and Gaussian process regression (Kalogirou, 2006;Zhao and Magoulès, 2012;Rastogi, 2016). Gaussian process regression has been proposed for optimisation (Wood et al, 2015;Wood, 2016;Gilan and Dilkina, 2015), optimal glazing design (Kim et al, 2013), model calibration (Monari, 2016;Heo and Zavala, 2012;Burkhart et al, 2014), and operational control (Yan et al, 2013). Most of these proposals have, however, remained theoretical.…”

Section: Introductionmentioning

confidence: 99%

Gaussian-Process-Based Emulators for Building Performance Simulation

Rastogi¹,

Khan²,

Andersen³

2017

Building Simulation Conference Proceedings

View full text Add to dashboard Cite

In this paper, we present an emulator of a buildingenergy performance simulator. Previous work on emulators for this application has largely focused on linear models. Since the simulator itself is a collection of differential equations, we expect non-linear models to be better emulators than linear models. The emulator we present in this paper is based on Gaussian-process (GP) regression models. We show that the proposed non-linear model is 3-4 times more accurate than linear models in predicting the energy outputs of the simulator. For energy outputs in the range 10-800 kWh/m 2 , our model achieves an average error of 10-25 kWh/m 2 compared to an average error of 30-100 kWh/m 2 from using linear models. In addition to being very accurate, our emulator also heavily reduces the computational burden for building designers who rely on simulators. By providing performance feedback for building designs very quickly (in just a few milliseconds), we expect our approach to be particularly useful for exercises that involve a large number of simulations, e.g., Uncertainty Analysis (UA), Sensitivity Analysis (SA), robust design, and optimisation.

show abstract

“…Bayesian calibration approaches have been successfully employed to integrate model-driven and data-driven procedures by training a Gaussian process meta-model with computer simulation data and using it in a Bayesian calibration process [8]. Monari et al introduced CALIBRO, an R package that has the objective of facilitating Bayesian calibration of BEMs [38]. Common approaches to perform optimization are reviewed in depth by Nguyen et al [39].…”

Section: Introductionmentioning

confidence: 99%

An Automated Building Energy Model Calibration Workflow to Improve Indoor Climate Controls

Yilmaz¹

View full text Add to dashboard Cite

White-box building energy models (BEM) are employed to optimize building operation and controls. Calibration improves the models' credibility by reducing the discrepancy between simulated and measured energy consumption data. However, calibrating a BEM is time-consuming and prohibitively expensive due to the large number of required model inputs and the limited availability of measurements. Moreover, this process remains challenging due to the lack of clear guidelines and a consensus on the calibration methodology. Therefore, the industry needs an efficient, low-cost, and accurate way to calibrate BEMs to actualize the benefits of operational optimization in commercial and industrial buildings. This research aims to increase the calibration efficiency by proposing an improved workflow to obtain a quick, lightweight, and parsimonious model that only requires building automation system (BAS) and energy meter data as well as simple geometric drawings. To this end, a deeper understanding of the uncertainty inherent in the BEM calibration process and the calibrated model's predictive performance on operational decisions were explored. The proposed method was demonstrated with a case study building in Ottawa, Canada, using metered energy use and actual meteorological year (AMY) data and operational parameters (e.g., HVAC system setpoints and schedules) extracted from the BAS. The results indicated more accurate parameters estimates and significantly more reliable energy consumption projections when the model is calibrated using energy meter data at a higher temporal resolution (i.e., hourly instead of monthly).Applying control interventions to calibrated case study BEM showed that up to 34% of energy could be saved through the optimized operation. Furthermore, leveraging BAS data not only overcame the overparameterization issue by reducing the number of unknown ii model inputs but was also found useful to detect operational anomalies to support the operational decision-making process.iii PrefaceThis integrated thesis consists of a conference paper that is published and a journal paper awaiting publication. Should readers wish to refer to materials from this thesis, the current thesis is required to be cited. The articles included in this thesis are as follow:

show abstract

CALIBRO: an R Package for the Automatic Calibration of Building Energy Simulation Models

Cited by 4 publications

References 17 publications

Increasing the efficiency of simulation-based design explorations via metamodelling

Increasing the efficiency of simulation-based design explorations via metamodelling

Gaussian-Process-Based Emulators for Building Performance Simulation

An Automated Building Energy Model Calibration Workflow to Improve Indoor Climate Controls

Contact Info

Product

Resources

About