Prediction of collapse potential via artificial neural networks

Collapse defined as the additional deformation of compacted soils when wetted is believed to be responsible for damage to the buildings resting on compacted fills as well as failure in embankments and earth dams. In this paper, three different types of neural networks, namely, conventional back-propagation neural network (BPNN), recurrent neural network (RNN), and generalized neural network (GRNN) are employed as computational tools to predict the amount of collapse and to investigate the influence of various parameters on the collapse potential. To arrive at this goal, 192 series of single oedometer test were carried out on three soils with different initial conditions and inundated at different applied pressures. The test results were used to prepare the necessary database for training the neural network. Similar test results available in literature were also included in the database to arrive at a total of 330 sets of data. Comparison of the network prediction for collapse potential with some available models shows the superiority of the network in terms of the accuracy of prediction. Moreover, relative importance of different factors on the collapse potential was assessed. Based on the neural network analysis, initial dry unit weight was found to be the most important factor influencing collapse potential.

[1]  G. David Garson,et al.  Interpreting neural-network connection weights , 1991 .

[2]  Dayakar Penumadu,et al.  Triaxial compression behavior of sand and gravel using artificial neural networks (ANN) , 1999 .

[3]  Jamshid Ghaboussi,et al.  New nested adaptive neural networks (NANN) for constitutive modeling , 1998 .

[4]  William N. Houston,et al.  Prediction of Field Collapse of Soils Due to Wetting , 1988 .

[5]  C. W. Lovell,et al.  Compressibility and Settlement of Compacted Fills , 1986 .

[6]  John Burland,et al.  Limitations to the Use of Effective Stresses in Partly Saturated Soils , 1962 .

[7]  A. McGown,et al.  THE COLLAPSE MECHANISM IN PARTLY SATURATED SOIL , 1973 .

[8]  Richard J. Fragaszy,et al.  REVIEW OF WETTING-INDUCED COLLAPSE IN COMPACTED SOIL. DISCUSSION AND CLOSURE , 1992 .

[9]  A. Goh.  Seismic liquefaction potential assessed by neural networks , 1994 .

[10]  Alan J. Lutenegger,et al.  DETERMINATION OF COLLAPSE POTENTIAL OF SOILS , 1988 .

[11]  G. A. Leonards,et al.  Reconsideration of Failure Initiating Mechanisms for Teton Dam , 1984 .

[12]  Deh-Shiu Hsu,et al.  BUILDING KBES FOR DIAGNOSING PC PILE WITH ARTIFICIAL NEURAL NETWORK , 1993 .

[13]  M. Mokhberi,et al.  A hyperbolic model for volume change behavior of collapsible soils , 1998 .

[14]  Musharraf Zaman,et al.  Modeling of soil behavior with a recurrent neural network , 1998 .

[15]  G. Habibagahi,et al.  A neural network framework for unsaturated soils , 2000 .

[16]  N. K. Bose,et al.  Neural Network Fundamentals with Graphs, Algorithms and Applications , 1995 .

[17]  P. Pells,et al.  THE FACTORS INFLUENCING COLLAPSE SETTLEMENT IN COMPACTED SOILS , 1975 .

[18]  Abu Kiefa.  GENERAL REGRESSION NEURAL NETWORK FOR PILES IN COHESIONLESS SOILS , 1998 .

[19]  Rui Zhao,et al.  Stress-Strain Modeling of Sands Using Artificial Neural Networks , 1995 .

[20]  E. Lawton,et al.  Collapse of compacted clayey sand , 1989 .

[21]  John H. Dudley.  REVIEW OF COLLAPSING SOILS , 1970 .

[22]  Adnan A. Basma,et al.  Evaluation and Control of Collapsible Soils , 1992 .

[23]  Anthony T. C. Goh,et al.  Neural-Network Modeling of CPT Seismic Liquefaction Data , 1996 .