Model Predictive Control of a Multilevel CHB STATCOM in Wind Farm Application Using Diophantine Equations

Mohammad Reza Nasiri, Shahrokh Farhangi, Jose Rodriguez

Research output: Contribution to journalArticlepeer-review

65 Citations (Scopus)


The finite control set model predictive control (FCS-MPC) for power electronic converters provides high dynamic performance, based on the limited number of inputs and accurate model of the converter. By applying this algorithm to multilevel converters such as a cascaded-H-bridge-based static var compensator (CHB STATCOM), the dynamic performance is degraded, because the optimized input is achieved by searching among a large set of switching combinations and redundancies. This paper proposes an FCS-MPC algorithm, which benefits high dynamic performance for the CHB STATCOM, despite the large set of inputs. The proposed FCS-MPC replaces the time-consuming optimization algorithm by solving Diophantine equations over the large set of switching combinations. The desired switching combination and all its redundancies are determined in a minimum execution time. The proposed FCS-MPC performance is validated by applying to two configurations: 1) a 15-level CHB STATCOM with energy storage capability for a short-term active power smoothing and reactive power compensation of a 10 MW fixed speed wind farm at medium voltage; and 2) an experimental seven-level CHB STATCOM at low voltage.

Original languageEnglish
Article number8353788
Pages (from-to)1213-1223
Number of pages11
JournalIEEE Transactions on Industrial Electronics
Issue number2
Publication statusPublished - 1 Feb 2019


  • Cascaded H-bridge based static var compensator (CHB STATCOM)
  • Diophantine equations
  • model predictive control (MPC)
  • wind farm

ASJC Scopus subject areas

  • Control and Systems Engineering
  • Electrical and Electronic Engineering


Dive into the research topics of 'Model Predictive Control of a Multilevel CHB STATCOM in Wind Farm Application Using Diophantine Equations'. Together they form a unique fingerprint.

Cite this