Prediction Method of TBM Key Tunneling Parameters Based on Real-time Operation Data

The optimal configuration of key tunnelling parameters of Tunnel Boring Machine (TBM) is the key to the safety and efficiency of tunnel construction. This paper proposes a method for predicting key tunnel parameters based on real-time TBM operation data to ensure real-time prediction and accuracy. The TBM working phases are first divided, then extract the important parameters affecting the prediction, and finally the key parameters are predicted using the Gated Recurrent Unit (GRU)neural network algorithm. The method was validated using TBM operation data from the Jilin Yin Song Project. The results shown that the method in this paper is highly accurate in predicting the three key parameters of total thrust, propulsion speed and cutter torque during the stable operation phase of the TBM. and a comparative study with three other algorithms proved that the chosen algorithm works best.