Identification of Models of Geomagnetic Activity and Space Weather Prediction

The problem of prediction of geomagnetic activity by the example of the method of Dst-index is considered. For its solving we suggested to use a nonlinear discrete dynamical system, which is obtained by decomposition of perturbations into series relative to correlation functions. Identification of structure and model parameters is defined by solving the corresponding problems of mathematical programming on bounded, but large enough set of stable models. On identification of model structure we took into account the assumption that model describes the state of mapetic-spherical-plasma medium, which is to great extent caused by influence of velocities of solar wind and south magnetic component of interplanetary magnetic field. The principal result of the work consists in numerical substantiation of existence of common stable model of minimal complexity, which provides the best prediction