V.A. Ivaniuk, V.A. Fedorchuk
The article is devoted to the problem of identifying models of nonlinear dynamic systems, presentedThe article is devoted to the problem of identifying models of nonlinear dynamic systems, presentedin the form of Volterra integral series. When applying the traditional approach, the keyproblem is to conduct numerous necessary experiments, which grows in an exponential relationshipwith respect to the dimension of the nucleus. A significant number of experiments, in manycases, make it impossible to use the Volterra series method for studying nonlinear dynamical systems.An adaptive method for identifying models of nonlinear dynamic systems is proposed,which allows obtaining models in the form of Volterra series with preservation of adequacy withrespect to models obtained by the traditional method, but the number of necessary experiments tobuild nuclei is less by about a one order of magnitude. The proposed method is considered toidentify thetwo-dimensional kernel. The results can be extended to cases with higher-orderVolterra kernels.
dentification of nonlinear dynamic systems, Volterra series, adaptive method ofdentification of nonlinear dynamic systems, Volterra series, adaptive method ofidentification.
2. Pupkov, K.A., Kapalin, V.I. and Yushchenko, A.S. (1976), Funktsionalnyye ryady v teorii nelineynykh sistem [Functional series in the theory of nonlinear systems], Nauka, Moscow, USSR.
3. Sidorov, D. (2013), Metody analiza integralnykh dinamicheskikh modeley: teoriya i prilozheniya [Methods of analysis of integral dynamic models: theory and applications], Izdatelstvo IGU, Irkutsk, Russia.
Full text: PDF