The special environment in the extra-long tunnels leads to safety accidents that are very easy to occur, and the traditional tunnel management mode has low degree of intelligence, which is not able to actively find the abnormal events in the tunnel and deal with them at the first time. For this reason, this paper proposes a kind of underground environment intelligent monitoring system, which is mainly applied to the real-time monitoring of tunnel bed settlement and deformation and gas concentration. In this system, the polar coordinate method of total station is used to obtain the three coordinates of different points, which is combined with multiple difference accuracy analysis to improve the accuracy of tunnel settlement and deformation monitoring. Based on the ARIMA model, a gas concentration prediction model is constructed, and different warning thresholds for gas concentration are set to realize the graded warning of gas in long tunnels. When monitoring the settlement and deformation of the roadbed, the deviation of the X, Y and H coordinates of each measuring station and the center error are within 1mm, and the difference between the automatic monitoring data and the manual review results fluctuates around ±0.01mm. When the ARIMA model is used to predict the gas concentration, the prediction error between the predicted value and the expected value fluctuates between ±0.05%, and a graded warning of gas concentration can be realized according to the confidence interval. The introduction of intelligent monitoring system in the underground environment of extralong tunnels helps to optimize the tunnel construction plan and ensure the safety of extra-long tunnel construction.
