On optimal input design in system identification for control

Abstract-- This paper considers a recently proposed framework for experiment design in system identification for control. We will consider model based control design methods, such as Model Predictive Control, where the model is obtained using a prediction error system identification method. The degradation in control performance due to uncertainty in the estimated model is specified by a constrained application cost function. The idea is to find a minimum power input signal, to be used in system identification experiment, such that the control performance specification is guaranteed with a given probability when using the estimated model. The objective is to provide insight in the potentials of this approach by using finite impulse response model examples, where it is possible to analytically solve the corresponding optimal input design problem. The examples show how the control specifications directly affects the excitation conditions in the system identification experiment.

• Type: Journal
• Year: 2010
• Where: CDC
• Authors: Bo Wahlberg, Håkan Hjalmarsson, Mariette Annergren
• Keywords: CDC 2010, Control Performance, Control Systems, Model, Model Predictive Control