Search

Building performance simulations (BPS) are used to test different designs and systems with the intention of reducing building costs and energy demand while ensuring a comfortable indoor climate. Unfortunately, software for BPS is computationally intensive. This makes it impractical to run thousands of simulations for sensitivity analysis and optimization. Worse yet, millions of simulations may be necessary for a thorough exploration of the high-dimensional design space formed by the many design parameters. This computational issue may be overcome by the creation of fast metamodels. In this paper, we aim to find suitable metamodeling techniques for diverse outputs from BPS. We consider five indicators of building performance and eight test problems for the comparison six popular metamodeling techniques – linear regression with ordinary least squares (OLS), random forest (RF), support vector regression (SVR), multivariate adaptive regression splines, Gaussian process regression (GPR), and neural network (NN). The methods are compared with respect to accuracy, efficiency, ease-of-use, robustness, and interpretability. To conduct a fair and in-depth comparison, a methodological approach is pursued using exhaustive grid searches for model selection assisted by sensitivity analysis. The comparison shows that GPR produces the most accurate metamodels, followed by NN and MARS. GPR is robust and easy to implement but becomes inefficient for large training sets compared to NN and MARS. A coefficient of determination, R 2 , larger than 0.9 have been obtained for the BPS outputs using between 128 and 1024 training points. In contrast, accurate metamodels with R 2 values larger than 0.99 can be achieved for all eight test problems using only 32–256 training points. [ABSTRACT FROM AUTHOR]

• Two systematic and one heuristic OAT approaches compared to Monte Carlo simulations. • 93.750 whole-year simulations of office room constitute the design space. • Bootstrapping used to consider numerous baselines, fixing orders and sample sizes. • Monte Carlo outperforms systematic OAT with a mere 32 random simulations. • Systematic OAT optimization finds local optima close to baseline – far from optimum. What is the best approach to perform building simulations as means to support decision-making and optimize building performance? Traditionally, the search for a compliant design is carried out in a manual one-at-a-time (OAT) approach where one design parameter is optimized before attention is shifted to the next one. In contrast, the Monte Carlo method makes it possible to combine many design parameters simultaneously and consider numerous input combinations. This paper presents a comparison of design approaches and their ability to optimize the performance of an office room with respect to energy demand and indoor climate. The test case is an office space with eight discretized inputs resulting in 93,750 possible designs. Bootstrapping is used to vary the baseline and fixing order for two systematic OAT approaches and to assess an increasing number of Monte Carlo samples. Summarily, OAT optimization depends heavily on the baseline and reveals only few, local optima close to this baseline, which are potentially far from the best solutions. On average, sampling 32 random simulations unveils better solutions and is less likely to suggest poor solutions. An increased sample of 1024 simulations reveals diverse solutions and favourable input ranges, while enabling sensitivity analysis, metamodeling, and flexible constraints. [ABSTRACT FROM AUTHOR]