Model Predictive Control of Modular Multilevel Converter

This chapter discusses how the predictive control often referred to as a model predictive control approach is becoming more popular and is widely employed to control the power converters. The cost function consists of multiple control objectives, which are optimized for weighting factors to reach the desired system behavior. The resulting switching state is then applied to the converter. The digital implementation of model predictive control (MPC) approach requires a discrete‐time model of the MMC system. The discrete‐time (DT) model is obtained from the discretization of continuous‐time (CT) model. The front‐end rectifier of MMC is modeled as a constant DC source. The rectifier output is connected to the positive and negative bars of each leg of an MMC system through a DC‐line. The DC‐line is modeled as an inductor and a series resistance. The output current flows through the converter legs to the load. The difference between the upper and lower arm current gives the magnitude of the current owing through the load. The circulating current flows among the converter legs due to the difference between the upper and lower arm voltages. The circulating current mainly consists of even‐order harmonic components corresponding to the multiples of twice the fundamental frequency.