Parametrical identification of linear dynamical system on the basis of stochastic difference equations

The method of parametrical identification of linear dynamic system is considered. This method uses results of measurements of instantaneous values of systems' impulse characteristics. The method is based on iteration procedure for mean-square estimation of coefficients of linear parametric discrete model in the form of stochastic difference equations. Results of numerically-analytical investigations and analysis of comparison with known methods for elimination of estimators' bias leads to the conclusion that suggested computational algorithm is effective and highly noise proof.