Abstract
A model of generalized regression neural network (GRNN) to evaluate collapsibility of loess is suggested in this paper, in which water content, saturation degree, dry density, void ratio and plastic index are taken as input neural cells and the output neural cell is coefficient of collapsibility. Selecting a series of distribution density of the radial basis function (spd), 20 group samples are tested after the GRNN is trained by 76 group samples, and the result is compared with experiment data from laboratory and the optimized smoothing parameter spd is obtained. The result shows that GRNN has a more satisfied prediction accuracy comparing with RBF if value of spd is adapted properly. GRNN is a new effective method for collapsibility assessment of loess.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Preview
Unable to display preview. Download preview PDF.
References
Ye, W.-m., Cui, Y.-j., Huang, Y., et al.: Collapsibility of loess and its discrimination criteria. Chinese Journal of Rock Mechanics and Engineering 25(3), 550–555 (2006)
Chen, X.-g., Pei, X.-d.: Technology and Application of Artificial Neural Network. China China Electric Power Press, Beijing (2003)
Li, X.-a., Peng, J.-b.: Estimation of loess slope stability with improved BP neural network. Journal of Geological Hazards and Environment Preservation 13(4), 56–59 (2002)
Chen, H.-j., Li, N.-h., Nie, D.-x., et al.: A model for prediction of rockburst by artificial neural network. Chinese Jounal of Geotechnical Engineering 24(2), 229–232 (2002)
Hikmet, K.C.: Application of generalized regression neural networks to intermittent flow forecasting and estimation. Journal of Hydrologic Engineering 10(4), 336–341 (2005)
Author information
Authors and Affiliations
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2011 Springer-Verlag Berlin Heidelberg
About this paper
Cite this paper
Zhang, Sh. (2011). Assessment of Loess Collapsibility with GRNN. In: Ma, M. (eds) Communication Systems and Information Technology. Lecture Notes in Electrical Engineering, vol 100. Springer, Berlin, Heidelberg. https://doi.org/10.1007/978-3-642-21762-3_97
Download citation
DOI: https://doi.org/10.1007/978-3-642-21762-3_97
Publisher Name: Springer, Berlin, Heidelberg
Print ISBN: 978-3-642-21761-6
Online ISBN: 978-3-642-21762-3
eBook Packages: EngineeringEngineering (R0)