Discrete time domain modeling and control of resonant converters

Abstract

Small-signal linearized modeling for resonant converters is of crucial improtance in many applications; it is not only improtant for assemssing stability and dynamic characteristics byt for designing compensators. However, the major part of modeling and analysis has, until now, concerned with a non-inner-feedback controlled resonant converter. Multi-loop control consisting of inner-feedback and outer voltage feedback look can provide excellent stability and dynamic charateristics for the converter. In the past, the inner-feedback loop has been developed intuitively because a systematic modelin technique can not be employed to design the inner-feedback loop. This thesis is concerned with modeling and control of resonant converters. A discrete time domain systematically applicable to an innerfeedback as well as a non-inner-feedback controlled resonant converter is proposed, and investigated through applichtions to resonant converters operating in combination with several types of control laws. In addition, several useful inner-feedback control laws are proposed, and design guidelines for these controls are presented based on the proposed modeling. In Chapter I, brief comments on the topics related to the switching converters are mode. Then the topics concerned with resonant converters are reviewed in detail, and the contents covered in this thesis are described. A discrete time domain modeling and analysis technique applicable to all types of inner-feedbacks as well as non-inner-feedback controlled series resonant converter(SRC) is presented in Chapter II. The nonlinear discrete time domain equations represention the static and dynamic behavior of the SRC are derived and linearized about the equilibrium state of the SRC. Also the inner-feedback control law is linearized about the equilibrium state. The linearized SRC and the linearized inner-feedback control law are then comnbined to arrive at a linearized inner-feedback controlled SRC. The linearized modeling is...