Abstract
This paper presents an Artificial Neural Network approach for a quick displacement prediction using the results of field measurement for NATM tunnels. The data obtained for NATM tunnels constructed in unconsolidated ground were analyzed with respect to key tunneling parameters, such as tunnel geometry, support condition and displacement observed during construction. Prediction of tunnel displacement at final stage was made at different timings of prediction. The results obtained suggest that the artificial neural network approach proposed here can predict tunnel deformations at the final stage with fairly high level of accuracy even for less information before construction.
