Modular Modeling of Complex Dynamics with Adams-Bashforth Sampled-Data Models for Control Purposes

JORDAN, MARIO; BUSTAMANTE, JORGE

Mar del Plata, Argentina

Simposio; Argentinean Symposium of Technology (AST’09); 2009

Sociedad Argentina de Informática (SADIO)

In this paper the features of discrete-time models in the form of interpolation approaches and black box models for control-system design purposes were compared. Their adequacy for modeling complex dynamics described as ordinary differential equations with some degree of interconnection were investigated. Particularly, two forms of modeling were developed, one following the traditional form with exact calculation of model coefficients and the other one employing system identification for such determination. As illustrated by means of a numeric example a comparison of features could be drawn out on the basis different structural parameters like the model order and the number of kernel. Finally a case study of a complex dynamics represented by an autonomous system in 6 degrees of freedom let check the goodness of the discrete-time models.