Abstract
A computer-aided method for the computation of optimal discrete-time models from continuous-time models is presented. A time-domain optimization technique is utilized for finding optimum values for the parameters of the discrete model by minimizing a criterion based on the squared error (MSE) between the time responses of the continuous-time model and the equivalent discrete-time model at sampling instants. Results of all examples considered show the strength of this technique as compared with other methods developed in the literature when one uses as the basis for comparison the MSE between step responses of the continuous model and those of the equivalent discrete models. Moreover, the proposed method gives accurate equivalent models for both small and large sampling periods.