Enhancing Multistep Finite Control Set Performance of 3L–NPC Converters using Optimal Pulse Patterns

This paper presents a model predictive control (MPC) strategy for the control and modulation of power converters. The method uses optimal pulse patterns (OPP) combined with a multistep finite–control–set (MFCS). It is well known that OPP can properly address the trade–off between the currents harmonic distortion and converter switching frequency. On the other hand, MFCS has already proved to be viable in real– time using commercial platforms such as field–programmable gate arrays. The paper leverages these features and proposes a framework where OPPs are used to generate current references tracked by an MFCS controller. A key attribute of the approach is the ability to generate these references online, storing the harmonic components of the OPP instead of its switching angles. A three–level neutral point clamped (3L–NPC) converter is presented as a case study. Results indicate a performance similar to that of OPP modulation alone but with a fast dynamic response, comparable to state–of–the–art MPCs.