A Robust Modified Model Predictive Control Algorithm for Quasi-Z Source Inverters

This paper presents a robust modified model predictive control (MPC) algorithm for quasi-Z-source inverters. In this algorithm, the predicted values of inductance current, capacitance voltage and load current of quasi-Z source network at the next sampling time are calculated based on the model. According to the multivariable control objective, the loss function is obtained and minimized, and the corresponding switch state is taken as the switch state at the next moment. It is worth noting that although the two null states have the same output voltage vector, they are often considered to be the same. However, the THD and junction temperature fluctuations produced by these two states are different, so the two different switching states are discussed separately in this paper. And the optimized switch vector can reduce the number of switches while maintaining good performance, thus reducing the thermal stress of power devices. In order to verify the proposed scheme, simulation and experiment are carried out.