Big Data-Based Performance Analysis of Tunnel Boring Machine Tunneling Using Deep Learning

In tunnel boring machine (TBM) construction, the advance rate is a crucial parameter that affects TBM driving efficiency, project schedule, and construction cost. During operation process, various types of indicators are monitored in real-time can help to control TBM. Although some studies have already been carried out prediction, research almost all based on statistical methods shallow learning algorithms, thereby having difficulties dealing with very large amount data modeling time-dependent characteristics parameters. To solve this problem, deep model proposed CNN architecture, bidirectional Long Short-Term Memory module, attention mechanism, which called CNN-Bi-LSTM-Attention model. first step, processed, architecture adopted extract features from sequence. Then Bi-LSTM module obtain indicators. The significant be addressed by added mechanism. training rotation speed cutter head (N), thrust (F), torque (T), penetration (P), chamber earth pressure (Soil_P) predict rate. influence periods performance also discussed. result shows not only amount, but an prediction. long-term may lead failure evaluation test cannot starting stage, denotes working state stage stable. Especially when starts work, prediction error big. compared several traditional methods, excellent