• DocumentCode
    3514078
  • Title

    StatCom´s B-Spline Neural Network Control

  • Author

    Ramirez, Juan M. ; Tapia, Rubén

  • Author_Institution
    CINVESTAV-IPN, Guadalajara
  • fYear
    2006
  • fDate
    15-18 Aug. 2006
  • Firstpage
    1
  • Lastpage
    6
  • Abstract
    The b-spline neural net is a convenient tool to execute the power system voltage control, with the possibility of carrying out such task on-line. The controller´s design simplicity and its performance compared with the conventional PI controller are shown, specially those related with the overshoots and control signals quality, impacting directly into the controlled variable response and having quieter behavior than that of the PI control. With this neural control the possibility to implement the on-line control is potential due to its relative simple and fast calculations, thus representing a great help for the control. This paper is aimed to exhibit the performance of the static compensator (StatCom) as a mean for controlling the power system voltage magnitude on those buses surrounding the neighborhood where the StatCom is embedded. Results show the technique´s possibilities on a multi-machine power system
  • Keywords
    PI control; control system analysis; control system synthesis; neurocontrollers; power system control; splines (mathematics); static VAr compensators; voltage control; StatCom; b-spline neural network control; conventional PI controller; multimachine power system; power system voltage control; static compensator; Automatic voltage control; Control systems; Neural networks; Pi control; Power system control; Power systems; STATCOM; Signal design; Spline; Voltage control; Neural net control; Power electronics; Voltage control;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    Transmission & Distribution Conference and Exposition: Latin America, 2006. TDC '06. IEEE/PES
  • Conference_Location
    Caracas
  • Print_ISBN
    1-4244-0287-5
  • Electronic_ISBN
    1-4244-0288-3
  • Type

    conf

  • DOI
    10.1109/TDCLA.2006.311456
  • Filename
    4104687