Prediction of TBM Advance Rate Considering Geotechnical and Operating Risks: An Example of the Lanzhou Long Water Conveyance Tunnel, China.

Highlights: A novel BAS-DNN model for prediction of TBM advance rate is developed. A database involving geotechnical and operating risks (RMR, TWCR, and RI) is set up. Quadratic relationships exist between RMR, TWCR, and AR. AR decreases while the descending rate gradually increases with RI increasing. The BAS-DNN model shows good performance in the tunnel section with a jamming event. [ABSTRACT FROM AUTHOR]