U.H. Jalolov, R.M. Bandishoeva, M.A. Badalova, Sh.Yo. Kholov

Annotation:

This article examines the problem of parametric identification of controlled dynamic systems during their normal operation using a model with adjustable parameters based on the integral modulation method. A key feature of this work is the adaptive nature of the identification procedure, which minimizes the negative impact of the controlled object's "swaying" when its parameters change. For many technological processes (TP) (catalyst aging in chemical engineering, wear of grinding media in grinding units, coating layer changes in rotary kilns), model parameters vary uncontrollably over time over a wide range.
This necessitates ongoing identification of control objects (CO) to restructure controller parameters. A promising solution to this problem is the development of reliable methods for automatically calculating the parameters of identification systems. Given current trends, such methods may rely on computer modeling and numerical optimization. This study is based on an identification method using adaptive models in a version that does not require controller shutdown. A signal processing approach based on wavelet-like finite functions, which are the kernels of the integration operator, is explored.

Keywords: Identification, integral modulation method, finite functions, parametric observer, modeling.