Geology prediction based on operation data of TBM: comparison between deep neural network and soft computing methods

Abstract

Tunnel boring machine (TBM) is a complex engineering system widely used for tunnel construction. In view of the complicated construction environments, it is necessary to predict geology conditions prior to excavation. In recent years, massive operation data of TBM has been recorded, and mining these data can provide important references and useful information for designers and operators of TBM. Therefore, a deep neural network-based geology prediction method is proposed in this work, which can provide relatively accurate geology prediction results according to the TBM operation data. The application case study on a tunnel in China shows that the proposed method can accurately estimate the geological conditions prior to excavation compared with the previous prediction methods based on soft computing methods. This work can be regarded as a good complement to the geophysical prospecting method during the construction of tunnels, and also highlights the applicability and potential of deep neural networks for other data mining tasks of TBMs.