System simplification using variable structure reduced models

Abstract

A practical method of model reduction for linear systems is proposed. To minimize the order of the reduced plant model and to achieve a very good approximation of the original system, the reduced model of the original system is constructed from two submodels. One model emphasizes the transient behaviour, and the other gives a very good steady-state approximation. The submodels are estimated using the techniques of frequency response matching and dominant energy modes analysis. The main features of the proposed method are that the order of the reduced model can be minimized and the reduced model can give good approximations in both the transient and steady-state responses of the original system. Moreover, applications of the proposed reduced model to control system design can lead to good performances in the transient and steady-state periods. Some simulated examples arc given to demonstrate the effectiveness of the proposed method.