Feng_Weixing.pdf (2.88 MB)
Download file

Space Vector Modulation for Three-Level Neutral Point Clamped Inverter

Download (2.88 MB)
posted on 24.05.2021, 10:46 by Weixing Feng
For most medium voltage high power applications, Neutral Point Clamped (NPC) three-level inverter is a preferred choice due to its advantages such as low cost, light weight and compact size. Space vector modulation is widely used in real-time digital control for power converters. It is especially suitable for use in NPC multilevel inverters due to its good harmonic profile, flexibility and easy digital implementation. This thesis focuses on the space vector modulation for high power three-level NPC inverter, where the switching frequency of the semiconductor devices is nominally below 1 OOOHz to reduce the power loss of the switching devices. The conventional space vector modulation (SVM) scheme for the three-level NPC inverter produces even order harmonics in the output voltages, which are not desirable for most industrial applications. In this thesis, the mechanism of even order harmonic generation is analysed. A new space vector modulation scheme, which can eliminate all the even order harmonics, is proposed. The performance of the new design is investigated and simulation results are provided for the verification purpose. The harmonic and THD profiles are compared with those of the conventional SVM scheme. The elimination of even order harmonics is achieved at the expense of a slight increase in the switching frequency. The proposed space vector modulation scheme can be applied to other types of converters for the even order harmonic elimination. An algorithm is developed to mitigate the neutral point potential deviation, which is a common problem in the NPC inverters. The simulation results show that this algorithm is suitable for both conventional and the proposed space vector modulation schemes.





Master of Engineering


Electrical and Computer Engineering

Granting Institution

Ryerson University

LAC Thesis Type


Thesis Advisor

Bin Wu

Usage metrics