A Novel VSC HVDC Frequency Control Strategy based on Neural Network Power Estimation using ROCOF

Soseul Jeong, Lee Junghun, Gilsoo Jang

    Research output: Contribution to journalConference articlepeer-review

    2 Citations (Scopus)

    Abstract

    This paper introduces a novel VSC HVDC frequency control strategy by Active Power Estimation. Although VSC HVDC has a much faster control speed than synchronous generator, frequency control of VSC is similar to that of governor. If VSC can use the fast-control character in frequency control, frequency stability can be improved. For fast control, proper power order is important and to estimate the power, this paper uses a neural network that solves the nonlinear relationship between input and output easily. To make quick control before the frequency reaches nadir, ROCOF is used as an input variable of the neural network. It can be seen that when the load of the system is greatly changed, the frequency fluctuation is significantly lowered when the VSC changes the output by the estimated power. Control of the VSC through the neural network is expected to enable faster frequency control than previously possible.

    Original languageEnglish
    Pages (from-to)467-471
    Number of pages5
    JournalIFAC-PapersOnLine
    Volume52
    Issue number4
    DOIs
    Publication statusPublished - 2019
    EventIFAC Workshop on Control of Smart Grid and Renewable Energy Systems, CSGRES 2019 - Jeju, Korea, Republic of
    Duration: 2019 Jun 102019 Jun 12

    Bibliographical note

    Publisher Copyright:
    © 2019, IFAC (International Federation of Automatic Control) Hosting by Elsevier Ltd. All rights reserved.IFAC (International Federation of Automatic Control) Hosting by Elsevier Ltd. All rights reserved.

    Keywords

    • Control
    • Frequency
    • Neural Network
    • Power Estimation
    • VSC HVDC

    ASJC Scopus subject areas

    • Control and Systems Engineering

    Fingerprint

    Dive into the research topics of 'A Novel VSC HVDC Frequency Control Strategy based on Neural Network Power Estimation using ROCOF'. Together they form a unique fingerprint.

    Cite this