Nonlinear System Identification with Shortage of Input-output Data
| Authors: | Sheng Feng, Beijing Institute of Technology, China
Chen J., Beijing Institute of Technology, China
Tu X.Y., Beijing Institute of Technology, China |
|---|
| Topic: | 1.1 Modelling, Identification & Signal Processing |
|---|
| Session: | Nonlinear System Identification II |
|---|
| Keywords: | System identification, Nonlinear system, ARMA models, Information integration, Parameter estimation |
|---|
Abstract
A system identification method for nonlinear systems with unknown structure by means of short input-output data is proposed. This method introduces more general model structure for nonlinear systems. Moreover, based on gray-box idea and its salient feature with expanding NARMAX (Nonlinear Autoregressive, Moving Average eXogenous) modeling, this method integrates different system information. Then GMDH (Group Method of Data Handling) method is employed to obtain the model terms and parameters. Effectiveness of the proposed method is illustrated by a typical nonlinear system with unknown structure and short input-output data.
