Order Reduction of Linear Dynamic Systems by Improved Routh Approximation Method

In this paper, a new simplified Routh approximation technique is proposed for the model order reduction (MOR) of large scale linear time-invariant systems. In reduced order modeling, the Routh approximation technique described in the literature is based on the alpha and beta parameters. This paper presents a new simplified Routh approximation technique for the MOR involving only alpha parameters to make the proposed technique simple. In this technique, the denominator polynomial of the lower order system is obtained by the Routh approximation technique and the numerator polynomial is computed by a simple mathematical algorithm as discussed in the proposed scenario. The additional advantage of the proposed method is that it always gives a stable reduced model if the original higher-order system is stable. To illustrate the proposed method, the fourth-order DC-DC converter model is reduced to its second-order reduced model. The modeling of DC-DC converter in continuous conduction mode is also developed and whose final output is a complete linear circuit model. In order to check the effectiveness and accuracy competitive to other popular and recent techniques in the literature, the proposed method has been applied on various standard numerical examples.