Open AccessOptimal finite state predictive direct torque control without weighting factors for motor drive applications
Author(s) -
IpoumNgome Paul Gistain,
MonNzongo Daniel L.,
SongManguelle Joseph,
Flesch Rodolfo C.C.,
Jin Tao
Publication year - 2019
Publication title -
iet power electronics
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.637
H-Index - 77
eISSN - 1755-4543
pISSN - 1755-4535
DOI - 10.1049/iet-pel.2018.5795
Subject(s) - control theory (sociology) , direct torque control , weighting , torque , model predictive control , computer science , power (physics) , optimal control , transient (computer programming) , automatic frequency control , induction motor , voltage , engineering , control (management) , mathematics , mathematical optimization , artificial intelligence , physics , acoustics , electrical engineering , thermodynamics , telecommunications , quantum mechanics , operating system
An optimal finite state predictive direct torque control without flux and switching frequency weighting factors (WFs) is proposed to avoid the tuning of WFs and to provide high dynamic performances with reduced switching frequency and computational burden. The proposed method is a combination of the well‐known direct torque control and multi‐objective ranking (MOR) priority methods. The classical MOR, used to avoid the flux WF, increases the computation time compared to the standard predictive torque control (PTC) and does not consider switching frequency issues. These drawbacks impose the need for high sampling rates, which increase the implementation cost and the inverter power losses, an important issue for medium and high‐power drives. In the proposed method, the direct torque control principle is used to reduce the number of voltage vectors from eight to three, which reduces the computational burden of MOR. Also, MOR is associated with a proposed priority scheme to reduce the switching frequency and avoid WFs. Simulation and experimental results confirm that similar steady‐state response as in the traditional PTC based on MOR is obtained. However, fast transient response is achieved, the computational burden is reduced by about 25%, and the switching frequency is substantially reduced compared to classical PTC.