Interactive Building Design Space Exploration Using Regionalized Sensitivity Analysis

Research output: Contribution to book/anthology/report/conference proceedingArticle in proceedingResearchpeer-review

58 Downloads (Pure)

Abstract

Monte Carlo simulations combined with regionalized sensitivity analysis provide the means to explore a vast, multivariate design space in building design. Typically, sensitivity analysis shows how the variability of model output relates to the uncertainties in models inputs. This reveals which simulation inputs are most important and which have negligible influence on the model output. Popular sensitivity methods include the Morris method, variance-based methods (e.g. Sobol’s), and regression methods (e.g. SRC). However, all these methods only address one output at a time, which makes it difficult prioritize and fixate inputs when considering multiple outputs. In this work, the primary outcome is a methodology to apply Kolmogorov-Smirnov two-sample (KS2) statistics to rank inputs due to their influence with respect to multiple outputs. A secondary outcome is the application of KS2 statistics in combination with the interactive parallel coordinate plot (PCP). The latter is an effective tool to explore stochastic simulations and to find high-performing building designs. The proposed methods help decision makers to focus their attention to the most important design parameters when exploring a multivariate design space. As case study, we consider building performance simulations of a 15.000 m² educational centre with respect to energy demand, thermal comfort, and daylight.
Original languageEnglish
Title of host publicationProceedings of the 15th IBPSA Conference : Building Simulation 2017
EditorsCharles S. Barnaby, Michael Wetter
Volume15
PublisherInternational Building Performance Simulation Association
Publication date2017
Pages726-735
ISBN (Electronic)978-1-7750520-0-5
DOIs
Publication statusPublished - 2017
EventThe 15th International Conference of IBPSA - Hyatt Regency San Francisco, 5 Embarcadero Center, San Francisco, CA 94111, San Francisco, United States
Duration: 7 Aug 20179 Aug 2017
Conference number: 15
http://www.buildingsimulation2017.org/

Conference

ConferenceThe 15th International Conference of IBPSA
Number15
LocationHyatt Regency San Francisco, 5 Embarcadero Center, San Francisco, CA 94111
CountryUnited States
CitySan Francisco
Period07/08/201709/08/2017
Internet address
SeriesBuilding Simulation Conference proceedings
Volume15
ISSN2522-2708

Fingerprint

Sensitivity analysis
Statistics
Thermal comfort

Keywords

  • Sensitivity Analysis
  • Regionalized Sensitivity Analysis
  • Kolmogorov-Smirnov
  • Parallel coordinate plot

Cite this

Østergård, T., Jensen, R. L., & Maagaard, S. (2017). Interactive Building Design Space Exploration Using Regionalized Sensitivity Analysis. In C. S. Barnaby, & M. Wetter (Eds.), Proceedings of the 15th IBPSA Conference: Building Simulation 2017 (Vol. 15, pp. 726-735). International Building Performance Simulation Association. Building Simulation Conference proceedings, Vol.. 15 https://doi.org/10.26868/25222708.2017.185
Østergård, Torben ; Jensen, Rasmus Lund ; Maagaard, Steffen. / Interactive Building Design Space Exploration Using Regionalized Sensitivity Analysis. Proceedings of the 15th IBPSA Conference: Building Simulation 2017. editor / Charles S. Barnaby ; Michael Wetter. Vol. 15 International Building Performance Simulation Association, 2017. pp. 726-735 (Building Simulation Conference proceedings, Vol. 15).
@inproceedings{edd1ae4ae717422d8e2081396fa8bd72,
title = "Interactive Building Design Space Exploration Using Regionalized Sensitivity Analysis",
abstract = "Monte Carlo simulations combined with regionalized sensitivity analysis provide the means to explore a vast, multivariate design space in building design. Typically, sensitivity analysis shows how the variability of model output relates to the uncertainties in models inputs. This reveals which simulation inputs are most important and which have negligible influence on the model output. Popular sensitivity methods include the Morris method, variance-based methods (e.g. Sobol’s), and regression methods (e.g. SRC). However, all these methods only address one output at a time, which makes it difficult prioritize and fixate inputs when considering multiple outputs. In this work, the primary outcome is a methodology to apply Kolmogorov-Smirnov two-sample (KS2) statistics to rank inputs due to their influence with respect to multiple outputs. A secondary outcome is the application of KS2 statistics in combination with the interactive parallel coordinate plot (PCP). The latter is an effective tool to explore stochastic simulations and to find high-performing building designs. The proposed methods help decision makers to focus their attention to the most important design parameters when exploring a multivariate design space. As case study, we consider building performance simulations of a 15.000 m² educational centre with respect to energy demand, thermal comfort, and daylight.",
keywords = "Sensitivity Analysis, Regionalized Sensitivity Analysis, Kolmogorov-Smirnov, Parallel coordinate plot",
author = "Torben {\O}sterg{\aa}rd and Jensen, {Rasmus Lund} and Steffen Maagaard",
year = "2017",
doi = "10.26868/25222708.2017.185",
language = "English",
volume = "15",
series = "Building Simulation Conference proceedings",
publisher = "International Building Performance Simulation Association",
pages = "726--735",
editor = "Barnaby, {Charles S.} and Michael Wetter",
booktitle = "Proceedings of the 15th IBPSA Conference",

}

Østergård, T, Jensen, RL & Maagaard, S 2017, Interactive Building Design Space Exploration Using Regionalized Sensitivity Analysis. in CS Barnaby & M Wetter (eds), Proceedings of the 15th IBPSA Conference: Building Simulation 2017. vol. 15, International Building Performance Simulation Association, Building Simulation Conference proceedings, vol. 15, pp. 726-735, The 15th International Conference of IBPSA, San Francisco, United States, 07/08/2017. https://doi.org/10.26868/25222708.2017.185

Interactive Building Design Space Exploration Using Regionalized Sensitivity Analysis. / Østergård, Torben; Jensen, Rasmus Lund; Maagaard, Steffen.

Proceedings of the 15th IBPSA Conference: Building Simulation 2017. ed. / Charles S. Barnaby; Michael Wetter. Vol. 15 International Building Performance Simulation Association, 2017. p. 726-735 (Building Simulation Conference proceedings, Vol. 15).

Research output: Contribution to book/anthology/report/conference proceedingArticle in proceedingResearchpeer-review

TY - GEN

T1 - Interactive Building Design Space Exploration Using Regionalized Sensitivity Analysis

AU - Østergård, Torben

AU - Jensen, Rasmus Lund

AU - Maagaard, Steffen

PY - 2017

Y1 - 2017

N2 - Monte Carlo simulations combined with regionalized sensitivity analysis provide the means to explore a vast, multivariate design space in building design. Typically, sensitivity analysis shows how the variability of model output relates to the uncertainties in models inputs. This reveals which simulation inputs are most important and which have negligible influence on the model output. Popular sensitivity methods include the Morris method, variance-based methods (e.g. Sobol’s), and regression methods (e.g. SRC). However, all these methods only address one output at a time, which makes it difficult prioritize and fixate inputs when considering multiple outputs. In this work, the primary outcome is a methodology to apply Kolmogorov-Smirnov two-sample (KS2) statistics to rank inputs due to their influence with respect to multiple outputs. A secondary outcome is the application of KS2 statistics in combination with the interactive parallel coordinate plot (PCP). The latter is an effective tool to explore stochastic simulations and to find high-performing building designs. The proposed methods help decision makers to focus their attention to the most important design parameters when exploring a multivariate design space. As case study, we consider building performance simulations of a 15.000 m² educational centre with respect to energy demand, thermal comfort, and daylight.

AB - Monte Carlo simulations combined with regionalized sensitivity analysis provide the means to explore a vast, multivariate design space in building design. Typically, sensitivity analysis shows how the variability of model output relates to the uncertainties in models inputs. This reveals which simulation inputs are most important and which have negligible influence on the model output. Popular sensitivity methods include the Morris method, variance-based methods (e.g. Sobol’s), and regression methods (e.g. SRC). However, all these methods only address one output at a time, which makes it difficult prioritize and fixate inputs when considering multiple outputs. In this work, the primary outcome is a methodology to apply Kolmogorov-Smirnov two-sample (KS2) statistics to rank inputs due to their influence with respect to multiple outputs. A secondary outcome is the application of KS2 statistics in combination with the interactive parallel coordinate plot (PCP). The latter is an effective tool to explore stochastic simulations and to find high-performing building designs. The proposed methods help decision makers to focus their attention to the most important design parameters when exploring a multivariate design space. As case study, we consider building performance simulations of a 15.000 m² educational centre with respect to energy demand, thermal comfort, and daylight.

KW - Sensitivity Analysis

KW - Regionalized Sensitivity Analysis

KW - Kolmogorov-Smirnov

KW - Parallel coordinate plot

U2 - 10.26868/25222708.2017.185

DO - 10.26868/25222708.2017.185

M3 - Article in proceeding

VL - 15

T3 - Building Simulation Conference proceedings

SP - 726

EP - 735

BT - Proceedings of the 15th IBPSA Conference

A2 - Barnaby, Charles S.

A2 - Wetter, Michael

PB - International Building Performance Simulation Association

ER -

Østergård T, Jensen RL, Maagaard S. Interactive Building Design Space Exploration Using Regionalized Sensitivity Analysis. In Barnaby CS, Wetter M, editors, Proceedings of the 15th IBPSA Conference: Building Simulation 2017. Vol. 15. International Building Performance Simulation Association. 2017. p. 726-735. (Building Simulation Conference proceedings, Vol. 15). https://doi.org/10.26868/25222708.2017.185