EASI RBD-FAST: an efficient method of global sensitivity analysis for present and future challenges in building performance simulation
| Autor: | Jeanne Goffart, Monika Woloszyn |
|---|---|
| Přispěvatelé: | Laboratoire Optimisation de la Conception et Ingénierie de l'Environnement (LOCIE), Université Savoie Mont Blanc (USMB [Université de Savoie] [Université de Chambéry])-Centre National de la Recherche Scientifique (CNRS), ADEME, OLIMP |
| Jazyk: | angličtina |
| Rok vydání: | 2021 |
| Předmět: |
Index (economics)
Computer science 020209 energy Computation 0211 other engineering and technologies 02 engineering and technology uncertainty and sensitivity analysis computer.software_genre 7. Clean energy Field (computer science) 021105 building & construction Architecture 0202 electrical engineering electronic engineering information engineering building energy modelling Morris method Sensitivity (control systems) Safety Risk Reliability and Quality Civil and Structural Engineering [STAT.AP]Statistics [stat]/Applications [stat.AP] ANOVA [SPI.GCIV.CD]Engineering Sciences [physics]/Civil Engineering/Construction durable time-dependent output Sobol sequence Building and Construction Variance (accounting) Mechanics of Materials Data mining computer Energy (signal processing) |
| Zdroj: | Journal of Building Engineering Journal of Building Engineering, Elsevier, In press, ⟨10.1016/j.jobe.2021.103129⟩ |
| ISSN: | 2352-7102 |
| DOI: | 10.1016/j.jobe.2021.103129⟩ |
| Popis: | International audience; In the field of building performance simulation, there is a growing interest for the use of sensitivity analysis (SA) and uncertainty analyses (UA), that enable to estimate the uncertainty in model prediction and to identify which model inputs are mainly responsible for this uncertainty. However, several methods exist, with different capabilities and limitations, leading to some misuse. The objective and the first novelty of the present study was to conduct a structured and comprehensive comparison of the capabilities of two different methods: EASI RBD-FAST (a variance-based computation of Sobol indices) and the popular Morris screening. This comparison was made on two annual outputs representative of winter (heating demand) and summer (overheating) from a comprehensive detached house energy model. It was shown that both methods fulfil basic expectations, since the same clusters of influential parameters were identified with the same computation effort. We have also shown that the EASI RBD-FAST method allows, with the same number of simulations, to extract very relevant additional information: intuitive sensitivity index and uncertainty of the output. Moreover, a novel application to compute temporal sensitivity indices on time-series outputs (such as free-floating temperature) was proposed. It demonstrated the possibility of using SA to investigate the dynamic properties of a building, such as persistence or inertia. In addition, the EASI RBD-FAST method illustrated in this paper proves to be not only very useful but also easy to use and accessible tool for building performance simulations. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|
Full Text from ScienceDirect