Shield Tail Seal Detection Method Based on Twin Simulation Model for Smart Shield

The tail sealing system of the shield machine is an important guarantee to ensure the tunneling of the shield machine underground. The safety warning of the shield tail sealing system is an important part of the intelligent shield machine, a prerequisite for the attitude adjustment of the shield machine and an important reference for the segment intelligent assembly robot. However, since the shield tail seal system works underground, it is difficult to construct experiments to verify the sealing performance and working state, so there is no mature Detection method for the shield tail seal. Therefore, this paper proposes a new detection method based on twin simulation-driven shield tail seal working status. First, a part of the working condition data of the existing construction site is selected as the training set of the simulation model to establish a twin simulation model, and then the reliability of the model is verified by using the verification set data. Then, based on this twin system, a large number of sample points are randomly selected for simulation to obtain corresponding data sets, so as to obtain the parameter range and corresponding relationship of various working states of the shield tail. Then according to the corresponding relationship between these data sets and states, a BP neural network detection and classification model is established. Finally, the twin simulation model is set to a new working condition, and the data generated by the simulation under this working condition is placed in the classification model to judge the working state, so as to verify the reliability of the detection model. The results showed that the detection accuracy was as high as 99.2%, which verified the reliability of the detection method. In short, the detection system has good stability and reliability, and meets the expected requirements of the design.