Sustainable Building Design Based on the Second Order Probability Approach

Abstract

Sustainable building design needs to consider uncertainty due to various sources, including weather condition, design uncertainty, and inherent random parameters, in order to maintain high performance. These different types of uncertainty requires more advanced computational methods. Therefore, this paper implements a second order probability approach to take into account two types of uncertainty: aleatory (inherent randomness) and epistemic (incomplete information). A retail building located in Tianjin (China) is used to demonstrate the application of this second order probability approach. The results indicate that this second order probability can deal with these two types of uncertainty in sustainable building design. This method can also provide insight on the dominance of design uncertainty or inherent aleatory uncertainty. Moreover, the variable importance method based on a locally weighted regression model is applied to determine the key variables influencing both aleatory and epistemic uncertainty of energy use for this retail building.