Abstract In designing a tunnel, numerical analyses play a very important role in assessing the influence of the excavation on the stability of the tunnel and the behavior of the surrounding ground. However, ground properties used for analysis include a lot of uncertainties because they are estimated through limited investigations and tests. For this reason, the behavior of a tunnel and the surrounding ground, and the rock class predicted at the design stage of the tunnel might be quite different from those measured during the tunnel excavation. In this study, a back analysis using not only tunnel measurement date but also face mapping data is performed to reduce the uncertainties included in ground properties. To this end, rock classes were inferred using face mapping data according to RMR (Rock Mass Rating) system and the possible ranges of ground properties were selected. They were utilized as bounds in constructing a training database for an artificial neural network analysis. To validate the proposed methodology, the results of this study were compared and analyzed with those of the previous studies.
A case study on the utilization of tunnel face mapping data for a back analysis based on artificial neural network
KSCE Journal of Civil Engineering ; 18 , 3 ; 751-759
2014-03-22
9 pages
Aufsatz (Zeitschrift)
Elektronische Ressource
Englisch
Network-level accident-mapping: Distance based pattern matching using artificial neural network
Online Contents | 2014
|Research on Speed Limit of Freeway Tunnel Group Based on Artificial Neural Network
British Library Conference Proceedings | 2011
|Test case based risk predictions using artificial neural network
Online Contents | 2006
|