Effect of Frequency Weights on Low Order Modeling of Linear Systems

Objectives: To develop a low order model of complex systems. Method: Modified balanced reduction with steady state gain factor to minimize the steady state error in the reduced order model. The performance of the model is assessed by calculating the Integral Square Error (ISE) and H-infinity norm. Findings: The paper presents a study to investigate the effect of adding input frequency weights in balanced reduction to obtain low order approximations of linear systems. To perform the study, systems belonging to various classes such as under-damped systems, critically-damped systems and over-damped systems are considered. Applications: Model order reduction has fundamental importance in control applications, for better understanding and simplified design of higher order systems.