A framework for performing uncertainty quantification is presented which is well-suited for systems with dependent inputs with unknown distributions. The multivariate input is given as a dataset whose variables can have strong, nonlinear dependencies. For each of the elements in the framework (dependency analysis, sample selection and sensitivity analysis), we recently developed new methods, which are here combined for the first time. The framework is tested on an example involving a wind farm simulation with offshore weather conditions as input.

Dependency analysis, Gaussian processes, Sensitivity analysis, UQ
European Conference on Computational Mechanics
Centrum Wiskunde & Informatica, Amsterdam, The Netherlands

Eggels, A.W. (Anne W.), & Crommelin, D.T. (2020). Uncertainty quantification with dependent inputs: Wind and waves. In Proceedings of the 6th European Conference on Computational Mechanics: Solids, Structures and Coupled Problems, ECCM 2018 and 7th European Conference on Computational Fluid Dynamics, ECFD 2018 (pp. 4099–4110).