ROBUST TECHNOLOGY FOR FORMING CORRELATION MATRICES OF MATHEMATICAL MODELS OF THE DYNAMICS OF CONTROL OBJECTS

Abstract

The authors analyze the challenges of forming correlation matrices in solving problems of identification of matrix models of the dynamics for real-life industrial objects. They propose generalized
algorithms that allow for reducing those matrices to similar matrices of useful signals. The specific
characteristics of real-life noisy technological parameters are taken into account and it is also
demonstrated that said algorithms can be used in the absence of a correlation between the useful
signal and the noise and in the presence of such correlation.
Keywords: stochastic process, identification, technological parameter, noise, noisy signal,
correlation function, correlation matrices, normalized estimates, dynamics models