Hybrid Modulated Model Predictive Control for Two Stage Matrix Converter with Reduced Computational Burden

The model predictive control (MPC) becomes more and more popular for power converters due to its advantages in recent years. For the Two Stage matrix converter (TSMC), the modulated model predictive control (M2PC) strategy not only keeps all the features of the traditional MPC, but also with the fixed switching frequency characteristic, which helps to improve the steady-state performance. However, the computational burden is heavy to execute the M2PC, which becomes a barrier to increase the sampling frequency for the control system. In this paper, a hybrid modulated model predictive control (HM2PC) strategy is proposed by combining a new vector selection method designed for the rectifier stage with a dead-beat control method designed for the inverter stage. Benefiting from this control architecture, the proposed strategy can effectively reduce the computational burden. Compared to the M2PC strategy, the execution time is considerably reduced to
$n$
without the control performance degradation. Moreover, the fixed switching frequency characteristic is also reserved. Comparative analysis and evaluation are carried out based on simulation and experimental results regarding the steady-state performance, dynamic performance, and execution time, which validate of the proposed strategy.