ADMM-Based Distributed Optimal Reactive Power Control for Loss Minimization of DFIG-Based Wind Farms

Sheng Huang, Peiyao Li, Qiuwei Wu, Fangxing Li, Fei Rong

Research output: Contribution to journalJournal articleResearchpeer-review


In this paper, a distributed optimal reactive power control (DORPC) scheme is proposed for minimizing the total losses of doubly fed induction generator (DFIG)-based wind farms (WFs), including the losses of generators, converters, filters, and net10 works. The DORPC minimizes total WF losses by optimally coordinating reactive power outputs of the DFIG stator and the grid-side converter. The optimal control problem is solved in a distributed manner by using the consensus alternating direction method of multipliers (ADMM). With the consensus ADMM, the total WF loss optimization problem is transformed into a dis13 tributed optimal power flow problem considered with DFIGs’ optimal operation. The optimization problem with local constraints considers the reactive power limit of DFIG-based wind turbines (WTs) and the voltage limits at all WT terminal buses inside the WF. In the DORPC, the optimal control problem is solved by the collector bus station controller and WT controllers in parallel, only with the information exchange between immediate neighbors. It eliminates the need of a central controller and centralized communication, implying better robustness and plug-and-play capability. A WF with 20 DFIG-based WTs was used to validate the proposed DORPC scheme.
Original languageEnglish
Article number105827
JournalInternational Journal of Electrical Power & Energy Systems
Number of pages18
Publication statusPublished - 2020


  • Alternating direction method of multipliers (ADMM)
  • Distributed reactive power control
  • Doubly fed induction generator (DFIG)
  • Loss minimization
  • Wind farm

Fingerprint Dive into the research topics of 'ADMM-Based Distributed Optimal Reactive Power Control for Loss Minimization of DFIG-Based Wind Farms'. Together they form a unique fingerprint.

Cite this