Application of Artificial Neural Networks to Seismic Waveform Inversion

  • Qiaodeng He
  • Hui Zhou
Part of the Modern Approaches in Geophysics book series (MAGE, volume 21)


A three-layer feedforward neural network is described, which has been applied to geophysical parameter inversion. The number of elements in the input layer is equal to the number of recorded samples. During training, the global minimum of the energy function is determined using a decimal-encoded genetic algorithm. It is necessary to use weights and thresholds of neurons as a gene group. To determine the weights and thresholds corresponding to the global minimum of the energy function, a large search range is used initially, which is then progressively reduced in order to accelerate convergence. The network was trained using both 5 and 10 numerically modeled records of the vertical component. After convergence, the network was tested using a randomly generated three-layer transverse isotropic model. The inversion results are very encouraging.


Genetic Algorithm Artificial Neural Network Hide Layer Energy Function Waveform Inversion 
These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.


Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.


  1. Calderon-Macias, C., and Sen., M. K., 1993, Geophysical interpretation by artificial neural systems: a feasibility study: 63`d Ann. Internat. Mtg., Soc. Expl. Geophys., Expanded Abstracts, 254257.Google Scholar
  2. Cheng, Z., and Zhu, G., 1995, Application of Kohonen network to the lateral prediction of oil and gas: Geophys. Prosp. Ptr., 34, 5356.Google Scholar
  3. Haykin, S., 1999, Neural networks — a comprehensive foundation: Prentice-Hall, Inc.Google Scholar
  4. Huang, K. Y., Liu, W. H., and Chang, I. C., 1989, Hopfield model of neural networks for detection of bright spots: 59th Ann. Internat. Mtg., Soc. Expl. Geophys., Expanded Abstracts, 444–446.Google Scholar
  5. Ji Fan, Fan Junbo, and Tan Yongdong, 1991, Neural network and application of neural computer: Westsouth Transportation University Publishing House.Google Scholar
  6. Liu, X., Xue, P., and Li, P., 1989, Neural network method for tracing seismic events: 59th Ann. Internat. Mtg., Soc. Expl. Geophys., Expanded Abstracts, 716–718.Google Scholar
  7. Michaels, P., 1994, Recurrent neural network representation of inelastic wave equation and full-waveform inversion without local minima: 64th Ann. Internat. Mtg., Soc. Expl. Geophys., Expanded Abstracts, 22–25.Google Scholar
  8. Murat, M. E., and Rudman, A. J., 1992, Automated first arrival picking: a neural network approach: Geophys. Prosp., 40, 587–604.Google Scholar
  9. McCormack, M. D., 1990, Seismic trace editing and first break picking using neural networks: 606 Ann. Internat. Mtg., Soc. Expl. Geophys., Expanded Abstracts, 321–324.Google Scholar
  10. Raiche, A., 1991, A pattern recognition approach to geophysical inversion using neural nets: Geophys. J. Internat., 105, 629–648.Google Scholar
  11. Roth, G., and Tarantola, A., 1994, Inversion of seismic waveforms using neural networks: 64`“ Ann. Internat. Mtg., Soc. Expl. Geophys., Expanded Abstracts, 788–791.Google Scholar
  12. Rothman, D. H., 1985, Nonlinear inversion, statistical mechanics and residual statics estimation: Geophysics, 50, 2784–2796.Google Scholar
  13. Sambridge, M. S., Tarantola, A., and Kennett, B. L. N., 1991, An alternative strategy for non-linear inversion of seismic waveforms: Geophys. Prosp., 39, 723–736.Google Scholar
  14. Veezhinathan, J., and Wagner, D., 1990, A neural network approach to first-break picking: Proc. Internat. Joint Conf. Neural Networks, San Diego CA, 1, 235–240.Google Scholar
  15. Zhou, H., and He, Q., 1995, Seismic wave modeling in heterogeneous medium by Hartley transform: Oil Geophys. Prosp., 30, 593–601.Google Scholar

Copyright information

© Springer Science+Business Media Dordrecht 2003

Authors and Affiliations

  • Qiaodeng He
    • 1
  • Hui Zhou
    • 1
  1. 1.Department of GeophysicsChangchun University of Science and TechnologyChangchunPeoples Republic of China

Personalised recommendations