A platform for research: civil engineering, architecture and urbanism
A platform for research: civil engineering, architecture and urbanism
An Intelligent Approach to Predict the Squeezing Severity and Tunnel Deformation in Squeezing Grounds
https://orcid.org/http://orcid.org/0000-0001-7301-5914
This study proposes a novel intelligent approach for predicting tunnel squeezing severity and tunnel deformation in squeezing grounds using two robust data mining techniques. This approach employs the logistic model tree (LMT) technique to predict the squeezing severity (LMT model) and the gene expression programming (GEP) technique to predict the tunnel deformation in squeezing grounds (GEP model). A database comprising the primary effective parameters on squeezing (i.e., tunnel depth (H), rock tunneling quality index (Q), support stiffness (K), and tunnel diameter (D)) that was assembled from 117 historical cases of 22 tunneling projects in different countries was used to train and test the models. The effectiveness of both models was evaluated using various performance metrics. The classification accuracy of the proposed LMT model on the test datasets was 0.882 and the coefficient of determination (R2) for the proposed GEP model on the test datasets was 0.817. These results prove high performance of proposed models for reliable prediction of tunnel squeezing and deformation. Furthermore, the proposed approach promotes transparency in the prediction process by providing white box models, which helps to take strategy regarding preventive measures such as optimization of tunnel rock support for controlling the surrounding rock deformation. Finally, the evaluation of importance for input variables revealed that K is the most contributing parameter for determining the tunnel deformation with a relative importance score of 0.43.
An Intelligent Approach to Predict the Squeezing Severity and Tunnel Deformation in Squeezing Grounds
https://orcid.org/http://orcid.org/0000-0001-7301-5914
This study proposes a novel intelligent approach for predicting tunnel squeezing severity and tunnel deformation in squeezing grounds using two robust data mining techniques. This approach employs the logistic model tree (LMT) technique to predict the squeezing severity (LMT model) and the gene expression programming (GEP) technique to predict the tunnel deformation in squeezing grounds (GEP model). A database comprising the primary effective parameters on squeezing (i.e., tunnel depth (H), rock tunneling quality index (Q), support stiffness (K), and tunnel diameter (D)) that was assembled from 117 historical cases of 22 tunneling projects in different countries was used to train and test the models. The effectiveness of both models was evaluated using various performance metrics. The classification accuracy of the proposed LMT model on the test datasets was 0.882 and the coefficient of determination (R2) for the proposed GEP model on the test datasets was 0.817. These results prove high performance of proposed models for reliable prediction of tunnel squeezing and deformation. Furthermore, the proposed approach promotes transparency in the prediction process by providing white box models, which helps to take strategy regarding preventive measures such as optimization of tunnel rock support for controlling the surrounding rock deformation. Finally, the evaluation of importance for input variables revealed that K is the most contributing parameter for determining the tunnel deformation with a relative importance score of 0.43.
An Intelligent Approach to Predict the Squeezing Severity and Tunnel Deformation in Squeezing Grounds
https://orcid.org/http://orcid.org/0000-0001-7301-5914
This study proposes a novel intelligent approach for predicting tunnel squeezing severity and tunnel deformation in squeezing grounds using two robust data mining techniques. This approach employs the logistic model tree (LMT) technique to predict the squeezing severity (LMT model) and the gene expression programming (GEP) technique to predict the tunnel deformation in squeezing grounds (GEP model). A database comprising the primary effective parameters on squeezing (i.e., tunnel depth (H), rock tunneling quality index (Q), support stiffness (K), and tunnel diameter (D)) that was assembled from 117 historical cases of 22 tunneling projects in different countries was used to train and test the models. The effectiveness of both models was evaluated using various performance metrics. The classification accuracy of the proposed LMT model on the test datasets was 0.882 and the coefficient of determination (R2) for the proposed GEP model on the test datasets was 0.817. These results prove high performance of proposed models for reliable prediction of tunnel squeezing and deformation. Furthermore, the proposed approach promotes transparency in the prediction process by providing white box models, which helps to take strategy regarding preventive measures such as optimization of tunnel rock support for controlling the surrounding rock deformation. Finally, the evaluation of importance for input variables revealed that K is the most contributing parameter for determining the tunnel deformation with a relative importance score of 0.43.
An Intelligent Approach to Predict the Squeezing Severity and Tunnel Deformation in Squeezing Grounds
Transp. Infrastruct. Geotech.
Ghasemi, Ebrahim (author) / Hassani, Saeed (author) / Kadkhodaei, Mohammad Hossein (author) / Bahri, Maziyar (author) / Romero-Hernandez, Rocio (author) / Mascort-Albea, Emilio J. (author)
Transportation Infrastructure Geotechnology ; 11 ; 3992-4016
2024-12-01
25 pages
Article (Journal)
Electronic Resource
English
Prediction of tunnel deformation in squeezing grounds
| British Library Online Contents | 2013
Prediction of tunnel deformation in squeezing grounds
| Elsevier | 2013
Prediction of tunnel deformation in squeezing grounds
| Online Contents | 2013