The derivation and implementation of the natural switching surfaces (NSS) considering certain parametric uncertainties for a flyback converter operating in the boundary conduction mode is the main focus of this paper. The NSS with nominal parameters presents many benefits for the control of nonlinear systems; for example, fast transient response under load-changing conditions. However, the performance worsens considerably when the converter actual parameters are different from the ones used in the design process. Therefore, a novel control strategy for NSS considering the effects of parameter uncertainties is proposed. This control law can estimate and adapt the control trajectories in a minimum number of switching cycles to obtain excellent performances even under extreme parameter uncertainties. The analytical derivation of the proposed adaptive switching surfaces is presented together with simulations and experimental results showing adequate performance under different tests, including comparisons with a standard PI controller.
Read full abstract