The Transactions of The Korean Institute of Electrical Engineers (전기학회논문지)

The Korean Institute of Electrical Engineers (대한전기학회)
권호 표지

A performance improvement of neural network for predicting defect size of steam generator tube using early stopping

조기학습정지를 이용한 원전 SG세관 결함크기 예측 신경회로망의 성능 향상

  • 조남훈 (숭실대 공대 전기공학부)
  • Published : 2008.11.01
Download PDF
⟨ Previous Next ⟩

Abstract

In this paper, we consider a performance improvement of neural network for predicting defect size of steam generator tube using early stopping. Usually, neural network is trained until MSE becomes less than a prescribed error goal. The smaller the error goal, the greater the prediction performance for the trained data. However, as the error goal is decreased, an over fitting is likely to start during supervised training of a neural network, which usually deteriorates the generalization performance. We propose that, for the prediction of an axisymmetric defect size, early stopping can be used to avoid the over-fitting. Through various experiments on the axisymmetric defect samples, we found that the difference bet ween the prediction error of neural network based on early stopping and that of ideal neural network is reasonably small. This indicates that the error goal used for neural network training for the prediction of defect size can be efficiently selected by early stopping.

Keywords

References

  1. D.P. Labridis, D.P. 'Comparative presentation of criteria for adaptive finite-element mesh generation in multiconductor eddy-current problems,' IEEE Trans. Magnetics, Vol. 36, No. 1, pp. 267-280, 2000 https://doi.org/10.1109/20.822537
  2. H.B. Lee, and D.H. Kim, 'Impedance calulation for a plate with crack in eddy current NDT using 3D indirect BIEM,' IEEE Trans. Magnetics, Vol. 36, No. 5, pp. 3131-3133, 2000 https://doi.org/10.1109/20.908709
  3. K. Miya, 'Recent advancement of electromagnetic nondestructive inspection technology in japan,' IEEE Trans. Magnetics, Vol. 38, No. 2, pp. 321-326, 2002 https://doi.org/10.1109/20.996088
  4. M. Rebican, Z. Chen, N. Yusa, L. Janousek, and K. Miya, 'Shape reconstruction of multiple cracks from ECT signals by means of a Stochastic method,' IEEE Trans. Magnetics, Vol. 42, No. 4, pp. 1079-1082, 2006 https://doi.org/10.1109/TMAG.2006.870967
  5. C. Gangzhu, A. Yamaguchi, and K. Miya, 'A novel signal processing technique for eddy-current testing of steam generator tubes,' IEEE Trans. Magnetics, Vol. 34, No. 3, pp. 642-648, 1998 https://doi.org/10.1109/20.668059
  6. H. Huang, T. Takagi, H. Fukutomi, 'Fast signal predictions of noised signals in eddy current testing,' IEEE Trans. Magnetics, Vol. 36, No. 4, pp. 1719-1723, 2000 https://doi.org/10.1109/20.877774
  7. S.J. Song and Y.K. Shin, 'Eddy current Flaw characterization in tubes by neural networks and finite element modeling,' NDT & E International, Vol. 33, pp. 233-243. 2000 https://doi.org/10.1016/S0963-8695(99)00046-8
  8. H. Haoyu, and T. Takagi, 'Inverse analyses for natural and multi cracks using signals from a differential transmit-receive ECT probe,' IEEE Trans. Magnetics, Vol. 38, No. 2, part 1, pp. 1009-1012. 2002 https://doi.org/10.1109/20.996259
  9. M. Rebican, N. Yusa, Z. Chen, K. Miya, T. Uchimoto, and T. Takagi, 'Reconstruction of multiple cracks in an ECT round-robin test,' International Journal of Appled Electromagnetics and Mechanics, Vol. 19, No. 1-4, pp. 399-404, 2004
  10. 조남훈, 이향범, 한기원, 송성진 '신경회로망을 이용한 원전SG 세관 결함패턴 분류성능 향상기법,' 전기학회 논문지, Vol. 56, No. 7, pp. 1224 - 1230, July, 2007
  11. M.T. Hagan, H.B. Demuth, and M.H. Beale, Neural Network Design, PWS Pub. Co, 1995
  12. S. Haykin, Neural Networks, New Jersey: Prentice-Hall, 1999
  13. L. Prechelt, 'Automatic early stopping using cross validation: quantifying the criteria,' Neural Networks, Vol. 11, No.4, pp. 761-767, 1998 https://doi.org/10.1016/S0893-6080(98)00010-0