ZEMCH 2015 - International Conference Proceedings | Page 86

In figure.6, regression model of different scenarios are shown. In this method one or more input
parameters are considered variable within their predicted variation range while others are kept
identical to their baseline values in all iterations. Coefficient of determination (R2) was considered
to study variance of obtained values via simulations that is explained by variation in expected
value, i.e. base case value. The low R2 value suggests that data is spread widely around the regression line while a large R2 indicate a strong relationship between obtained values and fitted model.

Figure 4: illustration of mean value and Standard deviation of 100 iterations in Monte Carlo simulations

Figure 5: Box plot of obtained results by 100 iteration of Monte Carlo simulation for 22 wall envelope
alternatives

84

ZEMCH 2015 | International Conference | Bari - Lecce, Italy