Computational & Technology Resources
an online resource for computational,
engineering & technology publications
PROCEEDINGS OF THE FOURTH INTERNATIONAL CONFERENCE ON SOFT COMPUTING TECHNOLOGY IN CIVIL, STRUCTURAL AND ENVIRONMENTAL ENGINEERING
Edited by: Y. Tsompanakis, J. Kruis and B.H.V. Topping
Hybrid RNN-GPOD Surrogate Model for Simulation and Monitoring Supported TBM Steering
S. Freitag, B.T. Cao, J. Ninic and G. Meschke
Institute for Structural Mechanics, Ruhr University Bochum, Germany
S. Freitag, B.T. Cao, J. Ninic, G. Meschke, "Hybrid RNN-GPOD Surrogate Model for Simulation and Monitoring Supported TBM Steering", in Y. Tsompanakis, J. Kruis, B.H.V. Topping, (Editors), "Proceedings of the Fourth International Conference on Soft Computing Technology in Civil, Structural and Environmental Engineering", Civil-Comp Press, Stirlingshire, UK, Paper 2, 2015. doi:10.4203/ccp.109.2
Keywords: surrogate models, recurrent neural networks, proper orthogonal decomposition, uncertainty, finite element, mechanised tunnelling, real time prognosis, monitoring.
In this paper, a hybrid surrogate modelling strategy for mechanised tunnelling simulations is presented. Recurrent neural network and gappy proper orthogonal decomposition approaches are combined to predict tunnelling induced time varying surface settlement fields. The hybrid surrogate model is trained and tested in the design stage of a tunnel project based on finite element analyses using a process-oriented simulation model to compute the surface settlements for selected scenarios of the tunnelling process parameters. During the construction, the surrogate model is used for real time predictions of the surface settlement to support the steering decisions of the machine driver. The surrogate model is continuously updated with the chosen process parameters for the steering of the tunnel boring machine and the corresponding surface settlement at selected monitoring points.
purchase the full-text of this paper (price £20)