• DocumentCode
    531952
  • Title

    System reliability calculation based on RBF neural network

  • Author

    Han, Yanbin ; Bai, Guangchen ; Li, Xiaoying

  • Author_Institution
    Sch. of Jet Propulsion, Beijing Univ. of Aeronaut. & Astronaut., Beijing, China
  • Volume
    1
  • fYear
    2010
  • fDate
    22-24 Oct. 2010
  • Abstract
    In order to improve the calculation accuracy of structural system failure probability, RBF neural networks which is based on the failure probability data calculated accurately can simulate into the four-dimensional space neural network of a single output neuron and three input neurons because system failure probability is related to reliability index, failure mode number and the correlation coefficient among the failure modes. As long as the networks meet the accuracy requirements and the training is completed, calculating neural network of system failure probability comes into being. The vector which is formed by the reliability index, failure mode number and the correlation coefficient is fed into the neural network, and then the probability of system failure can be accurately figured out. This method has high precision, the computational efficiency of fast and practical value in engineering calculations.
  • Keywords
    condition monitoring; probability; radial basis function networks; structural engineering computing; 4D space neural network; RBF neural networks; correlation coefficient; failure mode number; radial basis function networks; reliability index; structural system failure probability; Approximation methods; Artificial neural networks; Computer languages; Estimation; Neural network; failure mode; failure probability calculation; system reliability;
  • 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.2010.5619197
  • Filename
    5619197