Adaptive indirect model predictive control schemes for boost converters

Abstract

Boost converters are commonly used as the interface in renewable energy systems. To keep its output voltage at desired values, two adaptive Model Predictive Control (MPC) schemes with uncertainties in resistive load and external input voltage are proposed in this paper. Since the boost converter is characterized as a non-minimum phase system, controllers are not directly aimed at voltage regulation. The first method follows a cascade structure. A PI controller is introduced as the outer loop to generate a current reference, and MPC is designed as the inner loop by solving a finite horizon optimization of current tracking. Meanwhile, an unknown offset free MPC (UOFMPC) is proposed as the other scheme, where an observer is used for constructing an accurate model. Based on this exact model, MPC could realize voltage regulation accurately. Finally, both the proposed controllers are validated by simulation results from Matlab/Simulink.