Energy-Based Normalization for Resonant Power Converters

Resonant power conversion leads to advantages such as soft switching and high-frequency operation; however, analysis and design of a resonant converter require an efficient methodology. In general, conventional methodologies use a design-oriented architecture, which requires previously specifications and iterative circuit-level simulations. This procedure is the simplest; notwithstanding, it can be time consuming and inaccurate for more complex systems, as resonant converters. This paper proposes a result-oriented architecture that is independent of specifications; in other words, real circuit parameters are waived in exchange for generalized results. A normalization procedure is proposed by using an equivalence transformation to modeling the system with a state vector composed of terms as the square roots of the stored energy in the reactive components. The system is represented in a dimensionless state-space model, in which the terms of the state-space matrices are composed of resonant parameters, such as normalized resonant frequencies and quality factor; thereby, the system is independent of circuit parameters, such as inductances, capacitances, and resistances. The sixth-order hybrid Cuk converter is used to derivate the equations. Simulation results and an implementation of a 500-kHz hybrid buck converter are shown to compare theoretical, simulation, and experimental results.