The prediction by a mathematical model of the separation of uranium isotope
s using a gas centrifuge process is a hard task. The gas motion can be desc
ribed by analytical or numerical solutions of the system of equations defin
ed by the equation of continuity, the Navier-Stokes equation and the equati
on of energy. However, these calculations cannot be performed for actual ce
ntrifuges.
Neural networks are an alternative for modelling complex problems that show
too many difficulties to he solved by phenomenological models.
The authors propose the use of neural networks for the simulation and previ
sion of the separative and operational parameters of a gas centrifuge separ
ating uranium isotopes. The results from the uranium separation experiments
(Zippe data) are compiled and presented to the neural network in the learn
ing and testing processes. The prediction using the neural network model sh
ows good agreement with the experimental data.