Algorithms for improving adequacy of statistical identification

The possibility of providing adequacy of statistical identification by improving conditionality of correlation matrices is considered. Algorithms of calculation of the noise variance and microerrors are proposed. On the basis of these algorithms, the expressions for balancing the impact of microerrors in estimates of correlation functions are derived. It is shown that the values of relative errors of elements of the correlation matrices are equal as the result of balancing microerrors. This improves the conditionality of these elements and the results of the identification.