Recursive identification using feedforward neural networks. (English) Zbl 0830.93017

This paper considers input-output identification of nonlinear systems using neural network techniques. Using the concepts of generic observability and transversality, it is shown that a neural network can be used as an appropriate input-output identification model for a general class of nonlinear dynamical systems. The inputs to the neural network model are the most recent measurements of both the system’s inputs and its outputs. The approach exploits the approximation capability of the neural network and provide a theoretical basis for simulation results that illustrate the ability of a neural network to identify dynamical systems.


93B30 System identification
92B20 Neural networks for/in biological studies, artificial life and related topics
93C10 Nonlinear systems in control theory
Full Text: DOI