• DocumentCode
    2537921
  • Title

    Vectorial power flow calculation based on nonlinear programming

  • Author

    Yude, Yang ; Zhijun, Qin

  • Author_Institution
    Electr. Eng. Coll., Guangxi Univ., Nanning
  • fYear
    2008
  • fDate
    20-24 July 2008
  • Firstpage
    1
  • Lastpage
    6
  • Abstract
    For settle the divergence problem of ill-condition system, traditional power flow calculation in polar coordinates can be transformed into a nonlinear programming model. It is a convenient way to judge whether the general power flow equations have feasible solutions or not according to the value of target function. Getting the correction equations by Newton method, LDLT decomposition method is used to solve it. And the coefficient matrix of correct equation is reordered by using AMD reordering algorithm, thus the fill-in elements in LDLT decomposition is reduced significantly, which improves the calculating speed considerably. Proposed module and algorithm possess simplifying structure and can be implement or program easily with a novel vectorization expression. The adaptability and maintainability are also improved in this mode. Numerical simulations on test systems ranging in size from 118 to 1047 buses validate the correctness of the proposed model and method.
  • Keywords
    Newton method; load flow; matrix algebra; nonlinear programming; AMD reordering algorithm; Newton method; decomposition method; general power flow equations; matrix coefficient; nonlinear programming model; numerical simulations; polar coordinates; vectorial power flow calculation; Differential equations; Gaussian processes; Jacobian matrices; Load flow; Mathematical model; Matrix converters; Newton method; Nonlinear equations; Power system modeling; Symmetric matrices; Newton method; Nonlinear programming; Power flow calculation; Power system; Vectorization;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    Power and Energy Society General Meeting - Conversion and Delivery of Electrical Energy in the 21st Century, 2008 IEEE
  • Conference_Location
    Pittsburgh, PA
  • ISSN
    1932-5517
  • Print_ISBN
    978-1-4244-1905-0
  • Electronic_ISBN
    1932-5517
  • Type

    conf

  • DOI
    10.1109/PES.2008.4596447
  • Filename
    4596447