COMPUTER AIDED APPROACH FOR OBTAINING STABLE REDUCED ORDER MODELS

Abstract

The direct simulation or design in most of Large Scale Systems is neither computationally desirable nor physically possible.It is often desirable and even sometimes it becomes essential to reduce the order of the system for analysis and design. The main purpose of reducing the dimension of a large dynamic plant is to provide a simplified model which result in simple computation.Simplification techniques can be classified broadly into two main categories namely (i)Simplification in the frequency domain. (ii)Simplification in time domain. In this dissertation an attempt has been made to apply various order reduction techniques for linear dynamic systems in frequency domain and to compare the performance of original and reduced models.The performances of system model and reduced model have been compared to show that responses of reduced model are very close to the system model. iii