This paper is concerned with the convergence analysis of an identification algorithm for a class of nonlinear dynamic systems. The class includes systems that can be separated into symmetric zero-memory nonlinear gain parts and linear dynamic parts. A model reference adaptive technique is applied to the identification of such a nonlinear dynamic system after taking the following three points into consideration: 1) approximation of the nonlinear characteristics by a set of functions, 2) introduction of a set of parameters to obtain a multi-input, single-output system which is equivalent to the nonlinear system and 3) selection of an appropriate input signal. The convergence analysis of the proposed identifier is investigated from the stochastic view-point. A numerical example is presented to illustrate the effectiveness of the proposed identifier.

This content is only available via PDF.
You do not currently have access to this content.