Identification method based on Zadeh filter models

Abstract

Mathematical modeling which provides the description of objects and proper organization of control operations in future is an integral stage in the automation of production. One of the approaches to build a mathematical model of an object is to represent nonlinear systems as combinations of inertial and nonlinear inertialess elements. The models thus obtained are called block-oriented. In this paper, we consider nonlinear dynamic objects represented as the models of the Zadeh filter class. In the process of the method development the identification equations were derived for the case when the test signal is a single sinusoid. Then the case of two sinusoids was considered. Such investigations allowed us to identify the patterns and describe the general case for several test components in the signal. The results of digital modeling using the sum of harmonic signals confirm the feasibility and validity of the proposed approach for identifying nonlinear models of the Zadeh filter class.