• DocumentCode
    2794012
  • Title

    The study on self-adaptive predictive arithmetic based on RBF neural network applied in the proportion control of hydrogen and nitrogen in synthesis ammonia production

  • Author

    Hao, Lijun ; Wei, Xiaolei ; Wang, Zhihong ; Zhou, Shuai

  • Author_Institution
    Inst. of Electr. Eng. & Inf. Technol., Hebei Univ. of Sci. & Technol., Shijiazhuang, China
  • fYear
    2011
  • fDate
    15-17 July 2011
  • Firstpage
    881
  • Lastpage
    884
  • Abstract
    Aim at the control questions of more interference factors, time-variant and oversize time delay in the proportion control of hydrogen and nitrogen in synthesis ammonia production, a self-adaptive predictive PID control scheme based on RBF neural network theory is presented, using ahead predictive to overcome large delay, and PID arithmetic based on RBF network to adjust the parameter of controller on-line. Results of simulation experiment show that this method has quick system response, strong adaptability and better robustness, it will be has wide perspective and practicability for the proportion of hydrogen and nitrogen in synthesis ammonia production.
  • Keywords
    adaptive control; ammonia; chemical industry; chemical variables control; delays; hydrogen; neurocontrollers; nitrogen; radial basis function networks; stability; three-term control; NH3; RBF neural network; hydrogen; nitrogen; proportion control; robustness; self-adaptive predictive PID control; self-adaptive predictive arithmetic; synthesis ammonia production; time delay; Adaptive systems; Artificial neural networks; Nitrogen; Predictive models; Process control; Production; Radial basis function networks; RBF neural network; oversize time delay; proportion control of hydrogen and nitrogen; self-adaptive predictive PID control;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    Mechanic Automation and Control Engineering (MACE), 2011 Second International Conference on
  • Conference_Location
    Hohhot
  • Print_ISBN
    978-1-4244-9436-1
  • Type

    conf

  • DOI
    10.1109/MACE.2011.5987070
  • Filename
    5987070