Application of ANN for predicting pore water pressure response in a shake table test

Kumar, Bimlesh and Samui, Pijush (2008) Application of ANN for predicting pore water pressure response in a shake table test. In: International Journal of Geotechnical Engineering, 2 (2). pp. 153-160. (In Press)

PDF 10.3328.IJGE.2008.02.02.153-160.pdf Restricted to Registered users only Download (355kB) | Request a copy

Abstract

In this paper a proper assessment of pore water pressure has been done by using the shake table data. The pore water pressure ratio of a soil mass during an earthquake is dependent on both seismic and soil parameters. The impact of these soil variables on the pore water pressure ratio of soil is investigated through computational and knowledge based tools called neural networks. The artificial neural network is trained using actual shake table data. A forecast study is performed for pore water pressure in the shake table tank. The effect of different parameters such as frequency, no of cycles, void ratio, and depth of soil mass from ground level on pore water pressure ratio has been investigated. It has been found that the void ratio is having more effect on the pore water pressure while analyzing the shake table results.

Item Type:	Journal Article
Publication:	International Journal of Geotechnical Engineering
Publisher:	J. Ross Publishing, Inc.
Additional Information:	Copyright of this article belongs to J. Ross Publishing, Inc.
Keywords:	back-propagation;liquefaction;Levenberg-Marqudatt method;neural networks;pore water pressure;sensitivity analysis;void ratio
Department/Centre:	Division of Mechanical Sciences > Civil Engineering
Date Deposited:	04 Mar 2008
Last Modified:	19 Sep 2010 04:42
URI:	http://eprints.iisc.ac.in/id/eprint/13123

Actions (login required)

View Item