摘要
H~∞-IDENTIFICATIONOFINFINITE-DIMENSIONALLINEARSTOCHASTICSYSTEMSWEIChen;GUOLei(InstituteofSystemsScience,AcademiaSinicaBeijing?..
A method consisting of auto-regressive approximation in system identification and input normal realization truncation in model reduction is proposed for transfer function identification of infinite-dimensional linear stochastic systems. The identified low order transfer function is constructed on basis of the input-output data only, and the estimation error bounds in terms of L∞-norm are also derived.