Improved Deadbeat-Direct Torque and Flux Control for PMSM with Less Computation and Enhanced Robustness

Wusen Wang, Chunhua Liu, Hang Zhao, Zaixin Song

Research output: Journal article publicationJournal articleAcademic researchpeer-review

Abstract

Deadbeat-direct torque and flux control (DB-DTFC) significantly improves the torque response performance compared with the conventional direct torque control. However, the existing DB-DTFC algorithm suffers from a relatively heavy computational burden due to its intricate derivation. Thus, two simplified DB-DTFC algorithms for permanent magnet synchronous machines (PMSMs) are proposed in this article to relieve the burden. The first one is based on the stator flux differential, and the second one is based on the complex power of PMSM. Furthermore, the performance of DB-DTFC highly relies on the accuracy of machine parameters. Thus, modified stator flux observer and electromagnetic torque observer are designed to enhance parameter robustness based on disturbance observer theory. The theoretical derivation of the proposed methods is investigated in-depth. Finally, both simulation and experimental results verify that the proposed deadbeat methods can maintain the deadbeat performances with a reduced computational burden. Meanwhile, the modified observer can help the deadbeat algorithms work well when machine parameters change.

Original languageEnglish
Pages (from-to)1
Number of pages1
JournalIEEE Transactions on Industrial Electronics
DOIs
Publication statusAccepted/In press - 2022
Externally publishedYes

Keywords

  • Complex power
  • computational burden
  • deadbeat
  • direct torque control
  • disturbance observer
  • permanent magnet synchronous machine (PMSM)

ASJC Scopus subject areas

  • Control and Systems Engineering
  • Electrical and Electronic Engineering

Fingerprint

Dive into the research topics of 'Improved Deadbeat-Direct Torque and Flux Control for PMSM with Less Computation and Enhanced Robustness'. Together they form a unique fingerprint.

Cite this