Abstract
This paper outlines a new approach to the statistical identification and estimation of dynamic models for a widely applicable class of linear and nonlinear systems based on a combination of nonparametric and parametric estimation procedures. The new approach is applied to three systems: a chaotic version of the discrete-time logistic growth equation; a continuous-time, second-order nonlinear system with two feedback nonlinearities; and a predominantly linear but high-order, vibrating cantilever beam.
Keywords
Get full access to this article
View all access options for this article.
