The Research And Design Of Three-Phase Four-Leg Inverter Based On SVM

With the development of power electronics technology, a variety of unbalanced load areconnected to the grid, causing the grid voltage imbalance and power quality decline, seriouslyaffect the user’s normal use. Traditional three-leg inverter can output three-phase symmetricalvoltage in the case of three-phase symmetrical load. But it is difficult to output three-phasesymmetrical voltage when the three-phase load is asymmetric.In order to solve the impact ofunbalanced load on the power quality, the three-phase four-leg inverter which is obtained withthe addition of a bridge arm on the basis of traditional three-leg inverter is adopted in thispaper. Therefore, the three-phase symmetrical voltage can be acquired in the case ofunbalanced load, at the same time, with the advantages of simple structure, small size, lightweight, higher voltage utilization. The control strategy for three-phase four-leg inverter is asfollows:(1) The strategy based on three-dimensional space vector modulation is researched.Through the analysis of division of the24tetrahedrals,12tetrahedrals are invalid tetrahedrons,and the other12tetrahedrons are valid. The control effect of simplified algorithm isequivalent with the original algorithm. The original bloated and complex algorithms structurehas been simplified, becoming more compact and clear.(2) On the basis of simplified algorithm, the following two points are carried out:①Theswitching sequence of the traditional algorithm is9-stage, versus9-stage is adoped in thispaper. Therefore, the total number of switching times than the traditional algorithm can bereduced to1/4, and the switching losses is reduced25%.②Through the analysis of swithingsequence in the three-dimensional space, the cause of even harmonics is obtained. Under thepremise of meeting the requirement, by reasonable adjustment switching sequence, the outputwaveform can eliminate the even harmonics.(3) The algorithm is simulated using MATLAB/Simulink, and the simulation result isconsistent with the theoretical analysis. Finally, experiments which take the digital signalprocessor (TMS320F2812DSP) as control platform were carried out to verify theeffectiveness of the improved algorithm.