Reduced order modeling, time scale analysis, and simulation of power electronic systems

Switch mode operation has long been employed in power electronic systems due to the need for high efficiency. In recent years, this mode of operation has also been exploited in signal processing applications such as switched capacitor filters and switched capacitor radio frequency (RF) power amplifiers. While these systems have advantages in terms of weight, efficiency, and heat dissapation they are inherently highly nonlinear. This makes accurate analysis both challenging and computationally intensive. State-space based analysis is a powerful mathematical technique often employed for control system design, analysis, and implementation. These modeling tools are now gaining attention for the analysis of switch mode systems. This dissertation investigates the feasibility of applying a number of these techniques, including system balancing, time scale analysis, model order reduction, and working with the general solution to the state-space form to these nonlinear systems, for the purposes of design analysis, system minimization, and simulation. Two basic but dissimilar switched mode power converters are analyzed by employing a number techniques. The first of these systems is a common boost converter and the second is a Class E power converter that employs the technique of zero voltage switching in order to improve both efficiency and radiated RF emissions. In addition to the state-space based modeling techniques each power converter is also simulated using commercially available SPICE simulation tools. Finally, a rudimentary Class E power converter is constructed and tested in order to check the validity of these mathematical models.Ph.D.Includes bibliographical referencesby John Joseph McGarve