Finite control set model predictive control to guarantee stability and robustness for surface-mounted PM synchronous motors

Hoach The Nguyen, Jin Woo Jung

Research output: Contribution to journalArticlepeer-review

128 Scopus citations

Abstract

This paper proposes a finite control set model predictive control (FCS-MPC) to guarantee the stability and robustness for surface-mounted permanent magnet synchronous motor (SPMSM) drives. Continuous-input-based control laws are first developed from a control-Lyapunov function in order to both stabilize the closed-loop system via feedback control laws and ensure the robustness via online adaptive laws. Because the asymptotic stability of the proposed control method is guaranteed by at least one discrete switching-state, the continuous-input-based control laws are converted into relevant constraints of the FCS-MPC optimization problem. To validate the advantages of the proposed FCS-MPC, comparative studies with the conventional FCS-MPC and space vector modulation based adaptive control are conducted on a prototype SPMSM testbed with a TI TMS320F28335 DSP. The effectiveness of the proposed FCS-MPC is verified by the comparative schemes with/without the additional constraints. Superiority of the proposed schemes such as zero steady-state error, fast speed-tracking capability, well-regulated stator currents, and low average switching frequency are experimentally validated under the step-changes of load torque and speed reference.

Original languageEnglish
Pages (from-to)8510-8519
Number of pages10
JournalIEEE Transactions on Industrial Electronics
Volume65
Issue number11
DOIs
StatePublished - Nov 2018

Keywords

  • Adaptive control
  • control-Lyapunov function (CLF)
  • finite control set (FCS)
  • model predictive control (MPC)
  • permanent magnet synchronous motor (PMSM)

Fingerprint

Dive into the research topics of 'Finite control set model predictive control to guarantee stability and robustness for surface-mounted PM synchronous motors'. Together they form a unique fingerprint.

Cite this