Bitte wählen Sie ihr Lieferland und ihre Kundengruppe
The authors introduce the idea of using a criterion based on uncertainty propagation for appropriate detail for a model given a simulation task. A guideline to achieve this goal is proposed. A process to synthesize the differences in accuracy between the levels of the model of a component is developed so that the proposed guideline can be implemented in a solver. This approach comprises a preliminary simulation of the components, performed once for all, to determine the differences between model levels in a stochastic sense using the Collocation approach. The results of the preliminary analysis are then synthesized in a sensitivity function that we use in a rule-based selection of the levels for all the models in a simulation scheme. The efficacy of this method is demonstrated using case studies.