Time Invariant Modeling Of Higher Order DC-DC Converters

Abstract

A time invariant modeling technique is proposed for higher order dc-dc converters, including Cuk, SEPIC, and Zeta converters. This thesis also investigates some issues involved in applying existing averaging techniques such as circuit averaging, state space averaging, generalized state space averaging, Krylov-Bogoliubov-Mitropolsky method, and Floquet theory, to model the dynamic behaviors of higher order converters. The proposed methodology is shown to provide better accuracy compared to existing techniques, and distinguishes between different types of carrier signals of modulation, for instance, sawtooth and isosceles triangle carrier signal. This thesis offers a straightforward systematic procedure and summarizes this procedure with a list of rules, which makes the conversion from the state space model to the proposed time invariant model much easier. Hence, high order time invariant models can be easily developed for system design and stability analysis purposes. Both closed-loop time domain responses and open-loop frequency domain responses have been simulated and compared with the detailed switching model to verify the proposed methodology. Moreover, the simulation results of Cuk converter model are validated by experimental data for a 20W Cuk converter in both time and frequency domains.