• DocumentCode
    532176
  • Title

    Design and research for a multivariable neural network PID decoupling control algorithm with predictive compensation function

  • Author

    Pan, Hai-Peng ; Xu, Yu-Ying

  • Author_Institution
    Sch. of Mechanism & Autom., Zhejiang Sci. & Technol. Univ., Hangzhou, China
  • Volume
    7
  • fYear
    2010
  • fDate
    22-24 Oct. 2010
  • Abstract
    Analyzing and summarizing the advantages of the traditional decoupling algorithms, a new intelligent decoupling controller based on BP neural network PID with predictive compensation is designed to reduce the influence between the variables in multivariable, nonlinear and strong-coupling system. This kind of algorithm gains a decoupling performance by improving the structure of the traditional neural network PID algorithm. It also combines prediction control idea, which can accelerate the decoupling control speed and reduce the initial control overshoot effectively. Finally, the results of simulations using the mathematical model of air-cushioned headbox indicate that the algorithm has the characteristics of simple realization, quick dynamic behavior and small overshoot, which can be regarded as an innovation and improvement for the traditional decoupling algorithm.
  • Keywords
    backpropagation; compensation; intelligent control; multivariable control systems; neurocontrollers; nonlinear systems; predictive control; three-term control; BP neural network; air-cushioned headbox; multivariable neural network PID decoupling control algorithm; nonlinear system; predictive compensation function; strong-coupling system; Prediction algorithms; PID; air-cushioned headbox; decoupling control; neural network; predictive compensation;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    Computer Application and System Modeling (ICCASM), 2010 International Conference on
  • Conference_Location
    Taiyuan
  • Print_ISBN
    978-1-4244-7235-2
  • Electronic_ISBN
    978-1-4244-7237-6
  • Type

    conf

  • DOI
    10.1109/ICCASM.2010.5620093
  • Filename
    5620093