Computational & Technology Resources
an online resource for computational,
engineering & technology publications
Civil-Comp Proceedings
ISSN 1759-3433
CCP: 74
PROCEEDINGS OF THE SIXTH INTERNATIONAL CONFERENCE ON THE APPLICATION OF ARTIFICIAL INTELLIGENCE TO CIVIL AND STRUCTURAL ENGINEERING
Edited by: B.H.V. Topping and B. Kumar
Paper 30

Using BPNNs in selecting a Suitable Type of Foundation

H.K. Amin+, K.M. El Zahaby+, M.A. Taha* and A.S. Bazaraa*

+Housing and Building Research Center, Cairo, Egypt
*Faculty of Engineering, Cairo University, Egypt

Full Bibliographic Reference for this paper
H.K. Amin, K.M. El Zahaby, M.A. Taha, A.S. Bazaraa, "Using BPNNs in selecting a Suitable Type of Foundation", in B.H.V. Topping, B. Kumar, (Editors), "Proceedings of the Sixth International Conference on the Application of Artificial Intelligence to Civil and Structural Engineering", Civil-Comp Press, Stirlingshire, UK, Paper 30, 2001. doi:10.4203/ccp.74.30
Keywords: geotechnical, foundation, neural network.

Summary
In this paper, artificial neural networks have been used as tools to find a suitable foundation type for a given site with known characteristics.

Artificial neural networks are problem solving programs modeled on the structure of the human brain. Neural networks technology mimics the brain's own problem solving process. Just as humans apply knowledge gained from past experience to new problems or situations, a neural network makes use of previously solved examples to build a system of "neurons" that makes new decisions, classifications, and forecasts. Neural networks look for patterns in training sets of data, learn these patterns, and develop the ability to correctly classify new patterns or to make forecasts and predictions. Neural networks excel at problem diagnosis, decision making, prediction, and other classifying problems where pattern recognition is important and precise computational answers are not crucial.

Supervised backpropagation neural network (BPNN), the most widely used type of artificial neural networks (ANNs), has been used in this paper. A backpropagation network consists of many simple processing elements called neurons, grouped in layers and connections called synapses.

The "Brainmaker", a commercially available neural network simulator, has been used in the training of the neural network model. The training is accomplished by using 114 projects, 84 of which are collected from actual projects spread all over Egypt. Since the neural networks necessitate the existence of a significant number of projects that end up with each of the 4 used foundation types, and since it is practically difficult to collect such a number of projects in such a way to cover the required information for each of the 4 foundation types (isolated footings, strip footings, raft foundations and piled foundations), 30 additional projects have been generated making use of fuzzy set theory (FST). The details of using FST in selecting a suitable foundation type is explained and discussed in Paper 6.

A Comparative study has been accomplished to test the effect of the learning rates and the number of neurons on the number of misclassified projects.

A BPNN is thus built and is available for subsequent use by geotechnical engineers to select a suitable foundation type. Moreover, any additional data will help enhance the existing network.

purchase the full-text of this paper (price £20)

go to the previous paper
go to the next paper
return to the table of contents
return to the book description