Efficient Direct-Model Predictive Control with Discrete-Time Integral Action for PMSGs

Mohamed Abdelrahem, Christoph Hackl, Ralph Kennel, Jose Rodriguez

Research output: Contribution to journalArticle

11 Citations (Scopus)


This paper presents an efficient direct-model predictive control (EDMPC) technique for surface-mounted permanentmagnet synchronous generators (PMSGs) in variable-speed wind energy conversion systems (WECSs). The proposed control technique is based on directly computing the reference voltage vector (VV) from the demanded currents using a deadbeat (DB) function. Then, a discrete-time integral action (DTIA) is added to this DB function to enhance the robustness of the proposed EDMPC scheme against variations of the machine parameters and to achieve a good steady-state response. The proposed DTIA is simple and easy to implement. Finally, according to the location of this reference VV, two evaluations of the quality function are only required. Accordingly, the proposed EDMPC technique with DTIA overcomes the following drawbacks of the conventional direct-model predictive control (DMPC): i) High calculation burden, ii) sensitivity to parameters mismatches, and iii) non-zero steady-state error. The performance of the proposed EDMPC technique with DTIA has been experimentally investigated and compared with that of the EDMPC with time delay control approach (TDCA) and with the performance of the convention DMPC using a 14:5kW PMSG.

Original languageEnglish
JournalIEEE Transactions on Energy Conversion
Publication statusAccepted/In press - 1 Jan 2018


  • Adaptation models
  • Observers
  • parameter uncertainty
  • Permanent-magnet synchronous generators (PMSGs)
  • Predictive control
  • predictive control
  • Predictive models
  • Robustness
  • robustness
  • Steady-state
  • Switches
  • variable-speed wind turbines

ASJC Scopus subject areas

  • Energy Engineering and Power Technology
  • Electrical and Electronic Engineering

Fingerprint Dive into the research topics of 'Efficient Direct-Model Predictive Control with Discrete-Time Integral Action for PMSGs'. Together they form a unique fingerprint.

  • Cite this