PARALLEL GLOBAL OPTIMIZATION METHODS FOR IDENTIFICATION OF THE DYNAMIC BALANCE NORMATIVE MODEL OF REGIONAL ECONOMY

Constructed in Dorodnicyn Computing Center of the Russian Academy of Sciences dynamic models of regional economy contain a lot of unknown parameters. It is possible to identify them by minimization of nonlinear criteria of affinity for calculated and statistical data. In this paper a parallel index method of global optimization developed in Nizhny Novgorod State University is applied for that. The method uses a reduction of dimension on the basis of Peano curves and the information-statistical approach added with different updating.

Keywords

identification of mathematical model, regional economy, global optimization, index approach, local-global strategy, parallel algorithms.