STABLE OPTIMAL-MODEL REDUCTION OF LINEAR DISCRETE-TIME-SYSTEMS

A new algorithm is presented for finding reduced-order models of the l inear time-invariant systems. Proposed recursive algorithm minimizes a weighted mean squared impulse (or step) response error between the or iginal system and the reduced order model. This proposed optimization algorithm require solving very easy linear equations rather than the u sual complicated non-linear equations. The procedure guarantees a redu ced order stable model at each recursive stage. Method is easily exten ded to multi-input/multi-output systems. Algorithm's performance is co mpared with other methods reported in the literature.