• DocumentCode
    530621
  • Title

    A novel LS-SVM generalized inverse control of three-phase active power filter based on improved adaptive harmonic detection

  • Author

    Shao, Zhenhua ; Chen, Chong ; Huang, Yanwei

  • Author_Institution
    Dept. of Electr. Eng. & Autom., Fuzhou Univ., Fuzhou, China
  • Volume
    4
  • fYear
    2010
  • fDate
    24-26 Aug. 2010
  • Firstpage
    452
  • Lastpage
    455
  • Abstract
    Aiming at the problem that the traditional control has a limited control bandwidth and delays in digital implementation, a novel generalized inverse control of three-phase active power filter is studied in this paper. And the generalized inverse system can be approximated by LS-SVM due to the fast operational speed of LS-SVM. On the other hand, since the traditional algorithm has slow convergence rate and that the adaptive method with LPF has the drawback of slow dynamic response; an improved adaptive method for harmonic detecting based on integrated mean value is researched in this paper. In order to have a good linearization control effect, the generalized pseudo-linear system can be obtained by internal model and the generalized inverse system. At last, the simulation results show that a good steady-state performance and a rapid transient response can be obtained under the improved method.
  • Keywords
    active filters; harmonic analysis; least squares approximations; linearisation techniques; power harmonic filters; rectifying circuits; support vector machines; transient response; LPF; LS-SVM; adaptive harmonic detection; generalized inverse control; generalized pseudolinear system; three-phase active power filter; transient response; Indexes; Active Power Filter (APF); LS-SVM; adaptive harmonic detection; generalized inverse control; internal model control;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    Computer, Mechatronics, Control and Electronic Engineering (CMCE), 2010 International Conference on
  • Conference_Location
    Changchun
  • Print_ISBN
    978-1-4244-7957-3
  • Type

    conf

  • DOI
    10.1109/CMCE.2010.5610101
  • Filename
    5610101