Influence of Measurements Uncertainty on Uncertainty of Gilbert-Huang Transform Modes

The approach to simplifying the identification of a system using the input and output non-stationary or stationary processes is proposed in the article. The realizations of the process (input and output signals) are converted to the sum of individual components (modes) based on the Hilbert-Huang transform, and then the corresponding modes of this transformation are used for the virtual subsystem identification. The identification quality depends on the measurement uncertainty of both the signal and its Hilbert-Huang modes. The influence of measurements uncertainty of the signal on the individual modes' uncertainty, considering important factors, is analyzed in the article. The obtained results extend the conditions for solving the problems of system identification.