English Français

General overview

With the constant rising of computing power, the improvement of accuracy of both the numerical simulations and the experimental results to which they are compared, all this bring a new compulsory step in order to further build up analysis credibility : the uncertainty quantification ici.

This expression can have different meanings, depending on the targeted goals (sort out uncertain sources, reduce them, get a fair estimation of a critical probability or an optimal working point) and it can bring along mathematical models and methods (more or less complex).

The Uranie platform is a general open-source framework, developed at CEA (Commissariat à l'Energie Atomique et aux Energies Alternatives) in the energy division, in order to deal with uncertainty quantification, surrogate model generation, optimsations issues, code calibration, etc...

It benefits from both its dependencies and its own developpements, in order to provide an efficient data analysis framework, a on-the-flight C++ compiler and both a C++ and Python interface, advanced graphical tools and parallelisation options, etc...

The methods gathered in this tool are very generic and can then be applied to various field of research and applications (Uranie is using a black-box approach, always being non-intrusive with the code used).