A platform for research: civil engineering, architecture and urbanism
A platform for research: civil engineering, architecture and urbanism
Development of artificial neural networks and multiple regression models for the NATM tunnelling-induced settlement in Niayesh subway tunnel, Tehran
Abstract Here we investigate maximum settlement prediction of the Niayesh subway tunnel, excavated by employing the New Austrian Tunnelling Method in the Tehran metropolitan area, by several approaches, such as the semi-empirical method, linear and non-linear multiple regression method (MR), and finally by a programming Multi-Layered Perception (MLP) with a Back Propagation training algorithm. The geology at the site is mostly composed of conglomerates with pebbles and boulders. The maximum settlement is estimated based on the semi-empirical relations represented by several researchers. The input data set for MR and MLP models are soil characteristic [cohesion (C), internal friction angle (φ), elasticity modulus (E) and unit weight (Gs)], excavation depth (Z0), soil type (St) and PLAXIS 2D settlement prediction by the Hardening Soil model. Among all MLP and MR models, MLP models and especially model 6, the model based on E, Z, φ, Gs, C and St variables, seem to be reliable and agreeable to numerical results. The performance of MR, MLP, and optimized MLP models are evaluated by comparing statistic parameters, including coefficient correlations (R), root mean square error (RMSE), mean error (ME) and Akaike information criterion (AIC), whose values for model 6 are 0.93, 1.66, 0.89 and 13.16, respectively. Therefore, compared to other MLP and MR models, the optimized MLP model shows a relatively high level of accuracy. Additionally, model 4, the model based on E, Z, φ and Gs variables, shows in MLP analysis that unit weight does not have significant effect on maximum settlement.
Development of artificial neural networks and multiple regression models for the NATM tunnelling-induced settlement in Niayesh subway tunnel, Tehran
Abstract Here we investigate maximum settlement prediction of the Niayesh subway tunnel, excavated by employing the New Austrian Tunnelling Method in the Tehran metropolitan area, by several approaches, such as the semi-empirical method, linear and non-linear multiple regression method (MR), and finally by a programming Multi-Layered Perception (MLP) with a Back Propagation training algorithm. The geology at the site is mostly composed of conglomerates with pebbles and boulders. The maximum settlement is estimated based on the semi-empirical relations represented by several researchers. The input data set for MR and MLP models are soil characteristic [cohesion (C), internal friction angle (φ), elasticity modulus (E) and unit weight (Gs)], excavation depth (Z0), soil type (St) and PLAXIS 2D settlement prediction by the Hardening Soil model. Among all MLP and MR models, MLP models and especially model 6, the model based on E, Z, φ, Gs, C and St variables, seem to be reliable and agreeable to numerical results. The performance of MR, MLP, and optimized MLP models are evaluated by comparing statistic parameters, including coefficient correlations (R), root mean square error (RMSE), mean error (ME) and Akaike information criterion (AIC), whose values for model 6 are 0.93, 1.66, 0.89 and 13.16, respectively. Therefore, compared to other MLP and MR models, the optimized MLP model shows a relatively high level of accuracy. Additionally, model 4, the model based on E, Z, φ and Gs variables, shows in MLP analysis that unit weight does not have significant effect on maximum settlement.
Development of artificial neural networks and multiple regression models for the NATM tunnelling-induced settlement in Niayesh subway tunnel, Tehran
Mohammadi, S. D. (author) / Naseri, F. (author) / Alipoor, S. (author)
2014
Article (Journal)
Electronic Resource
English
BKL:
56.00$jBauwesen: Allgemeines
/
38.58
Geomechanik
/
38.58$jGeomechanik
/
56.20
Ingenieurgeologie, Bodenmechanik
/
56.00
Bauwesen: Allgemeines
/
56.20$jIngenieurgeologie$jBodenmechanik
RVK:
ELIB18
Bored tunnelling - NATM tunnelling
| British Library Conference Proceedings | 2000
Kramer Tunnel—Challenging NATM Tunnelling in Germany's Alps
| British Library Conference Proceedings | 2021
Artificial neural networks analysis of Sao Paulo subway tunnel settlement data
| British Library Online Contents | 2008