Generalized Model Predictive Pulse Pattern Control Based on Small-Signal Modeling—Part 1: Algorithm

A model predictive controller based on optimized pulse patterns is proposed that is suitable for higher-order linear systems, such as converters with $LC$ filters. The controller manipulates the switching times of an optimized pulse pattern. The switching time modifications are approximated by the strengths of impulses, which are based on a small-signal linearization around the nominal switching instants. With these, the evolution of the state variables over the prediction horizon can be described by a set of linear differential equations. An objective function penalizes the predicted tracking error of the controlled variables, such as the converter currents, filter capacitor voltages, and grid currents, over a prediction horizon. Thanks to the use of impulse strengths, the underlying optimization problem is a convex quadratic program, which can be solved in real time to determine the switching time modifications of the pulse pattern to be applied by the controller.