Model predictive current control for multilevel CHB-PMSM system with lower calculation

摘要

Due to the large number of voltage vectors in the multilevel inverter, the traditional multilevel model predictive control has a problem of heavy online computing burden. In this study, a model predictive current control strategy for multilevel-cascaded H-bridge inverter-permanent-magnet synchronous motor (CHB-PMSM) system is proposed. By deeply analysing the relationship between current vector and increment of the voltage vector, the candidate voltage vector sets in dynamic and steady state are optimised. Besides, the number of candidate voltage vectors is reduced to two regardless of the levels of CHB inverter, avoiding all voltage vectors participating in the calculation. Consequently, the experimental results on the CHB-PMSM system verify that the proposed strategy can significantly reduce the computational complexity and make the motor system obtain good dynamic and steady-state performance.