On Model Reduction of Multi-Input-Multi-Output Discrete Interval Systems

Analysis and design of higher order systems are complex tasks involving high computational burden. So, these problems should be tackled using model order reduction technique. This paper deals with an efficient model order reduction technique for multi-input-multi-output (MIMO) discrete interval system. For the proposed system, the denominator of different transfer functions of lower order MIMO discrete interval model (LOMDIM) is computed by retaining dominant poles of higher order MIMO discrete interval system (HOMDIS) and the numerators of LOMDIM are computed by matching alpha coefficients. This technique is verified with the help of two-input-two-output (TITO) test system.