Formulation of multi-hazard damage prediction (MhDP) model for tunnelling projects in earthquake and landslide-prone regions: A novel approach with artificial neural networking (ANN)

The two most common natural disasters in the Himalayas are earthquakes and landslides. Disaster-proof auditing is required for ongoing transportation infrastructure projects in this region. The Artificial Neural Networking (ANN) approach is used in the present study to train neural networks with input layers in terms of disaster parameters, structural configuration, and confining medium characteristics. This model will aid in predicting tunnel failure damage states during earthquake and landslide events. The proposed damage indices for various damage states of the portal and lining can be applied to define the co-seismic demand for individual structural elements. Co-seismic design recommendations will be useful in determining transportation infrastructure serviceability in post-disaster conditions.