Decentralized Volt/Var Control Based on Variable Gradient Projection for PMSG-based Wind Farm

Hanzhi Peng, Sheng Huang, Qiuwei Wu, Feifan Shen, Wu Liao, Xueping Li

    Research output: Contribution to journalJournal articleResearchpeer-review

    Abstract

    This paper proposes a decentralized Volt/Var control method based on variable gradient projection (VGP) for the permanent magnet synchronous generator (PMSG)-based wind farm (WF). The aim is to achieve a fast Var response of each wind turbine (WT) to depress voltage fluctuations inside the WF caused by the stochastic wind and voltage disturbance of external girds. The gradient projection (GP) method is used in the PMSG-based WF to regulate bus voltage within a feasible range in a fully decentralized manner while the dynamic characteristics of the PMSGs and grid side converters (GSCs) are taken into consideration. Moreover, according to the difference of the available power distribution among the WTs inside the WF, the adaptive and variable gain method is designed for the GP method, which can achieve maximum reactive power margin (RPM) of each WT during the real-time operation of the WF. With the proposed control method, the PMSG-based WTs are coordinated to achieve a near global optimal performance only with local measurements. The voltage fluctuations of buses are reduced while the RPM of WT converters is maximized. The WF with MW WTs is tested in Matlab/simulink with the real-time full nonlinear AC power flow environment.
    Original languageEnglish
    JournalIEEE Transactions on Sustainable Energy
    Volume13
    Issue number3
    Pages (from-to)1305-1314
    ISSN1949-3029
    DOIs
    Publication statusPublished - 2022

    Keywords

    • Decentralized control
    • Reactive power margin
    • Voltage fluctuations
    • Variable gradient projection method
    • Wind farm

    Fingerprint

    Dive into the research topics of 'Decentralized Volt/Var Control Based on Variable Gradient Projection for PMSG-based Wind Farm'. Together they form a unique fingerprint.

    Cite this