Design of identification algorithms with the covariance matrix unknown

A plant is discussed that is described by a mathematical model having a known vector of input variables. Changes in the output variables are recorded by using multichannel measurement of indirect indices of the plant output. It is required to develop an algorithm for estimating the unknown parameters of the plant model with incomplete a priori information on the statistical characteristics of the output variable measurement channels. A three stage procedure is suggested for design of identification algorithms. The same procedure leads to new identification algorithms.