• DocumentCode
    477479
  • Title

    NdFeB Magnet Composite Design Based on BP Network

  • Author

    Huiyu Wang ; Yuanhua Ren ; Jianfeng Pan

  • Author_Institution
    Coll. of Comput. Sci. & Technol., Zhejiang Univ., Hangzhou
  • Volume
    1
  • fYear
    2008
  • fDate
    20-22 Oct. 2008
  • Firstpage
    204
  • Lastpage
    208
  • Abstract
    Traditionally, the composite design of NdFeB magnet was consuming with repeated experiments. In order to speed up the design process of NdFeB magnet, BP network is trained to model the relationship between NdFeB magnet composition and the magnetic properties. The training data that come from production data are classified into two groups for the purpose of simplifying the network structure and improving the efficiency of the training. The experiment results of regression analysis indicate that the predicted data are consistent with the measured data quite well. The tendencies of the magnetic properties are also forecasted in the experiment results, which will guide the NdFeB magnet composite designing.
  • Keywords
    backpropagation; boron alloys; data handling; iron alloys; materials science computing; neodymium alloys; neural nets; pattern classification; permanent magnets; regression analysis; BP network; NdFeB; NdFeB magnet composite design; magnetic properties; network structure; regression analysis; training data; Artificial neural networks; Biological system modeling; Intelligent networks; Magnetic properties; Neural networks; Neurons; Predictive models; Process design; Production; Training data; BP Network; NdFeB magnet design; properties forecasting; training data classify;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    Intelligent Computation Technology and Automation (ICICTA), 2008 International Conference on
  • Conference_Location
    Hunan
  • Print_ISBN
    978-0-7695-3357-5
  • Type

    conf

  • DOI
    10.1109/ICICTA.2008.160
  • Filename
    4659473