Downsampled Iterative Learning Controller for Flyback CCM Inverter

Abstract

We propose a downsampled iterative learning controller (ILC) for a flyback inverter operating in continuous conduction mode (CCM). The flyback CCM inverter features step-up/-down ability, small number of circuit components, and high conversion efficiency, so it is well suited to the distributed renewable energy systems. But the conventional proportional-integral (PI) controller for the flyback CCM inverter shows poor steady-state response because the inverter itself suffers from time-varying grid-voltage disturbances and its transfer function has a right-half plane zero. The ILC itself achieves acceptable steady-state response of the flyback CCM inverter but can excite large overshoot in the iteration domain; this overshoot may destroy the solid state devices in the power circuit. To overcome this problem, we use a downsampled ILC technique that guarantees monotonic convergence of tracking error. The proposed ILC is computationally simple and easy to implement. Numerical simulations and experimental tests validate the proposed control approach.