Identification of Nonlinear Systems Using Dynamical Neural Networks: A Singular Perturbation Approach
In this paper, we employ singular perturbation analysis to examine the stability and robustness properties of a dynamical neural network identifier. Various cases that lead to modeling errors are taken into consideration. Input-output stability theory is used to assure convergence and internal stability of the identifier. Not all the plant states are assumed to be available for measurment.
KeywordsSingular Perturbation Reduce Order Model Robustness Property Adaptive Observer Unknown Plant
Unable to display preview. Download preview PDF.
- H. Nijmeijer and A.J. van der Schaft, Nonlinear Dynamical Control Systems, New York, NY, Springer-Verlag, 1989.Google Scholar
- K.S. Narendra, Stable Identification Schemes in System Identification: Advances and Cases Studies, New York: Academic, 1976.Google Scholar