Neural network model for dynamic displacement of twin water tunnels
Neural network model for dynamic displacement of twin water tunnels
Fatemeh Madadi1 HamidReza VosoughiFar2
1) College of Engineering, South Tehran branch , Islamic Azad University , Tehran ,Iran
2) College of Engineering, South Tehran branch , Islamic Azad University , Tehran ,Iran
IOI Code
:
XYCZ-AAGFB
Publication :
International Conference on Engineering and Applied Science(iceasconf.com)
Abstract :
Proper estimation of dynamic displacement of the seismic behavior of tunnels and they have always been of interest to researchers. Where tunnels are built to be water, it is more important use of virtual networks and their relationship with numerical methods can provide a proper analysis in connection with seismic them. Therefore, in this study the effects of three earthquake near and far fields for twin water tunnels is studied. The twin water tunnels of Siahbishe dam located in Mazandaran, Iran was selected as case study. Twin tunnels using advanced finite element software taking into account the time history is analyzed and the internal pressure of the water in the tunnels. Mohr-Coulomb failure criterion for analytic been modified and Riley damping criteria is on action. Based on dynamic analysis tunnels, the peak of displacement was observed when model was subjected to near field earthquake. Next, numerical results processed by the virtual network to estimate a link were trained by incorporating a code designed by the authors (ANN) and the best-trained network that could express the relationship between inputs and outputs with minimal error rate and maximal regression was obtained (R = 0.96 and MSE = 8.95E-4)
Keywords : time history dynamic analysis twin tunnel neural network