An Identification By Adapting The Fictitious Controller To Data
In the present paper, we propose a new identification method by adapting the ``fictitious controller" to the obtained data in the closed loop experiment.
A fictitious controller, which is introduced in this paper, consists of the nominal model with unknown plant parameters and the implemented controller used in the closed loop.
One of the key points of the present method is the adaptation of such a controller to the actual experiment data.
Moreover, the adaptation can be performed by using an off-line nonlinear optimization.
Since the required material in the proposed method is only one-shot experimental data under the normal operation like step responses, this method has a practical advantage in the sense that the costs and time for the identification can be reduced.