摘要:In the present work a methodology to perform uncertainty, sensitivity and importance analysis of large numerical models is presented. This methodology combines several advanced statistical tools. The proposed tools applied are independent of the model under analysis and the model itself is treated as a black box. This flexibility allows for the analysis of a variety of numerical models. As an example it can be mentioned the application to finite element models, boundary element models, logical models (boolean equations), etc.
其他摘要:In the present work a methodology to perform uncertainty, sensitivity and importance analysis of large numerical models is presented. This methodology combines several advanced statistical tools. The proposed tools applied are independent of the model under analysis and the model itself is treated as a black box. This flexibility allows for the analysis of a variety of numerical models. As an example it can be mentioned the application to finite element models, boundary element models, logical models (boolean equations), etc.