COMPARING TWO METHODS OF STOCHASTIC MODELING FOR BUILDINGS

In buildings research Monte Carlo (MC) methods are used for uncertainty analyses. This paper presents an alternative approach for uncertainty analysis in buildings, coined the internal method. Here, the joint probability density functions (pdfs) of model outputs are directly computed from pdfs of the inputs. This method needs only a single simulation run to compute the output pdfs compared to significantly higher number of runs for MC methods. The application of the two competing techniques (MC and internal method) to the most elemental building problem (1R1C) reveals that MC methods prove to be much more appropriate for building problems.