IDENTIFICATION OF CONTINUOUS-TIME SYSTEMS FROM SAMPLES OF INPUT-OUTPUT DATA USING THE DELTA-OPERATOR

The use of the delta-operator for estimating the parameters of a discrete-time model for a continuous-time system is investigated. Algorithms are derived for estimating these parameters from the samples of the input-output data either for a transfer function model or for a state-space model. Results of simulation indicate that the algorithms work very well even if the sampling frequency is ten times the largest undamped natural frequency of the system to be identified, whereas severe numerical difficulties arise with models based on the unit advance (q) operator normally associated with discrete-time models at such high sampling rates.