Identification of Nonlinear Systems by Using Elliptic and Loxodromic Methods

Abstract

This article describes the results of using elliptic and loxodromic methods for identification of nonlinear systems with applying of the distributed computing and Acsocad software. Presents the existing methods for identification of objects and systems in the time domain, and also describes standard methods and identification algorithms by using Acsocad. The research of nonlinear objects with two and three unknown parameters are done. For their identification the methods of global minimum search by ellipse and spiral for two-dimensional case and with using ellipsoid and loxodromic curves for three-dimensional case are developed. This article describes the results of researches, which contain the parameters estimation error, standard deviation, the number of performed operations, and also the map of the global minimum search.