Prediction of pile settlement based on cone penetration test results: An ANN approach

Several theoretical and experimental methods are available to estimate pile settlement. Due to difficulties for obtaining undisturbed samples, many of these methods have been focused on in-situ tests. The cone penetration test is one of the most effective in-situ tests because of its geometrical ana...

Full description

Saved in:
Bibliographic Details
Published in:KSCE journal of civil engineering 2015, 19(1), , pp.98-106
Main Authors: Baziar, Mohammad Hassan, Saeedi Azizkandi, Alireza, Kashkooli, Armin
Format: Article
Language:English
Subjects:
Analogies
Artificial neural networks
Civil Engineering
Cone penetration tests
Consistency
Engineering
Experimental methods
Field tests
Geotechnical Engineering
Geotechnical Engineering & Applied Earth Sciences
Industrial Pollution Prevention
Learning theory
Mathematical models
Network analysis
Neural networks
Parameters
Penetration
Pile settlement
Piles
Predictions
Samples
토목공학
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Several theoretical and experimental methods are available to estimate pile settlement. Due to difficulties for obtaining undisturbed samples, many of these methods have been focused on in-situ tests. The cone penetration test is one of the most effective in-situ tests because of its geometrical analogy with the piles as well as presenting continuous results along the depth. In this study, 1300 recorded settlement data from 101 pile loading tests with the CPT results were collected. Then Artificial Neural Network analyses (ANN) were conducted to obtain the best model for the prediction of pile settlement. The relative importance of input parameters has been evaluated using senility analysis. Accuracy predictions of the proposed model, along with other classic methods, were compared with the recorded values from the loading tests with the aid of different statistical parameters. This comparison indicated the superiority of the proposed model over previous methods. A parametric study has also been performed for the input parameters to study the consistency of the suggested model.
ISSN:1226-7988
1976-3808
DOI:10.1007/s12205-012-0628-3