A new identification approach with application to uncertainty assessment for H-Infinity control
DEI - Room 3B
November 24th, 2010
The goal of this research is that of setting up a new system identification approach, called "min-max identification", which effectively exploit both a-priori and a-posteriori (the data) information on the system.
The identification problem is formulated as a chance-constrained optimization program and a randomized solution with guaranteed accuracy is provided by using the so called "scenario approach".
The min-max approach to identification is then used to derive a description of the uncertainty on the unknown data generating system which is suitable for H-infinity control.