Objective function at optimization of separation cascades

Abstract

The optimization of the parameters for separation cascades is of great importance in their designing. The optimal cascade is usually construed as an installation which allows obtaining the required amount of a marketable product at the lowest unit cost. In many situations pertaining to the case, the cascade comprising a minimum possible number of separation elements (for example, gas centrifuges) has the minimum unit cost. However, there are other criteria, optimization by which leads to another set of the cascade parameters. In this connection it seems important to analyze the objective functions used at the optimization as well as their interactions. The paper proposes a method of optimizing the cascades for the isotope separation based on genetic algorithms. The use of such optimization methods in the theory of cascades is promising because they are resistant to specifying initial approximations and practically independent of the objective function type, as well as easily adaptable to changes in the dimension of the space. These specific features are often the main problems in optimizing the parameters of separation cascades. As an example, optimizing the separation cascade with variable stage separation factors for natural uranium enrichment is considered. The optimization is performed by various criteria of efficiency. The results obtained have made it possible to compare the parameters of the cascades optimized by various criteria and to identify the relationships between them. The developed method can be adapted to optimizing various cascade options for the separation of both binary mixtures and multi-component mixtures.