
Comparative study of three model predictive current control methods with two vectors for three‐phase DC/AC VSIs
Author(s) -
Park SoYoung,
Kwak Sangshin
Publication year - 2017
Publication title -
iet electric power applications
Language(s) - English
Resource type - Journals
ISSN - 1751-8679
DOI - 10.1049/iet-epa.2016.0687
Subject(s) - current (fluid) , control theory (sociology) , model predictive control , three phase , phase (matter) , control (management) , materials science , computer science , engineering , physics , voltage , electrical engineering , artificial intelligence , quantum mechanics
Three model predictive current control methods based on two‐vector utilisation are proposed and compared voltage source inverters (VSIs) to regulate the load currents and achieve a low current ripple and error. In the three methods, selecting and dividing two voltage vectors for a next sampling interval are determined by an optimisation. However, the procedure of determining two vectors and their durations differ for each method. The first method considers all 49 possible combinations obtained by two vectors. In the second method, the prior vector is selected beforehand using the conventional method, and then, the latter vector is selected among seven vectors along with durations. The second scheme has the advantage of lower computational complexity; however, it suffers from an increase in current error, relative to the first scheme. The least computationally complex approach is achieved using the third method choosing two vectors in the same manner as the conventional method but suffering from the largest current errors among the proposed methods. As a result of the optimisation procedure involved in selecting and partitioning two vectors, all three proposed methods result in a reduced total harmonic distortion and smaller current errors in comparison to the conventional method with half sampling period.