Model order reduction and approximation analysis for control system design

Dynamic systems are mostly complex, time-varying and of higher order. This paper presents model order reduction scheme used for the order reduction of complex, higher order dynamic systems. The proposed scheme for reduction of higher order system transfer function is based on modified-pole clustering and coefficient matching technique. The efforts has been made in this paper to decrease the order of the given transfer function and essentially preserving all the useful characteristics of the original transfer function using the proposed scheme. Some other techniques used for the reduction of model dimension are also described and analyzed. An example has been included to illustrate the proposed method of model order reduction for analysis and control of a dynamic system. The present method is compared with other order reduction method and concluding remarks are presented.