Artificial Neural Network-Based Method for Locating The Latest Assembled Shield Segment Ring

The shield segments need to fit the design tunnel axis (DTA) during construction, and the axis fitting deviation needs to be maintained within a permissible range. The location of the latest assembled segment ring is an indispensable condition for calculating the axis fitting deviation of new ring. However, there is no description of the calculation method for the location of the latest assembled segment ring. Aiming at this problem, an artificial neural network-based method for calculating the location of the ring is proposed. First, the calculation of the segment ring location is transformed into the calculation of the relative location of the ring to the shield machine, and the thrust cylinder strokes and shield tail gaps are used to calculate the relative position. Then, a Monte Carlo-based method is proposed to generate dataset, which is used to train neural network to obtain the calculation model of relative location. Finally, the calculation method of segment ring location was tested on the test set, and the results showed that the calculation results meet the engineering requirements. The proposed method can accurately calculate the location of the latest assembled segment ring in real time, which can improve the quality of tunnels.