| Abstract |
The Neural Point Clamped (NPC) converter has found a wide range of application in MV, high power drives. Minimization of losses and regulation of Neutral Point Voltage (NPV) of this converter are the objects aimed in this paper. The converter losses and variations of NPV are measured from the simulation results for possible switching patterns of Space Vector Modulation (SVM). IGBTs with linear fall and tail current model are selected as power switches of converter. Through a comparative study among the converter losses and variations of NPV resulted from the all-possible switching sequences of three level SVM, it is shown that a unique switching algorithm cannot minimize both the switching losses and variations of NPV. Therefore the paper proposes a sub-optimized switching algorithm resulting in very satisfactory conditions for the converter from the viewpoint of losses and of variations of NPV. Using this switching algorithm, the converter losses and variations of NPV reach either to their minimum or to a value near their minimum. |
