Fast Transient Response of Series Resonant Converters Using Average Geometric Control

Linear controllers and small-signal modeling techniques have been used for over 40 years to control power electronic converters around the equilibrium operating point. Outside the quiescent area, the shortcomings of linear controllers are well known and include uncertain large-signal transient behavior and sluggish dynamic/recovery response. In particular, resonant converters have inherent large-signal behavior, making the transient performance of small-signal linear controllers poor. This paper develops a geometric controller based on large-signal modeling to achieve enhanced dynamic response of the series resonant converter (SRC). The proposed modeling strategy provides detailed insight into the behavior of the SRC and creates an analytical framework to successfully perform geometric control. A new geometric control law is mathematically derived using average circular trajectories producing accurate and fast dynamics during startup, sudden load or reference changes. In order to validate the theoretical analysis, experimental and simulation results of an SRC are presented and compared to those of a linear and a nonlinear controller. The dynamic response is, on average, over 100 times faster than that of the designed linear controller, while overshoot is virtually eliminated.