The adaptive identification algorithms of dynamical systems based on the principle of regularity

Abstract

The questions of the adaptive algorithms synthesis for identifying dynamic systems based on the principle of regularity are considered. Estimates of the error in setting the initial data for various model structures, which allow, without making a direct decision, to estimate the error of the desired solution from above and choose the optimal value of the regularization parameter, are obtained. Based on methods for solving incorrectly posed problems, regular recurrent iterative computational schemes for solving the parametric identification problem are obtained. Keywords control object; control system; parametric identification; dynamic filtering; regularization; regularization parameter