The selection of material parameters relates to the excavation stability of the underground caverns. However, back analysis is an efficient method to evaluate mechanical parameters. Given the defects of BP neural network, such as low capability of generalization and long training time, by using GA, which have global optimization ability to optimize the BP neural network weights. The parameter of surrounding rock was designed by uniform and orthogonal method, not only reduced the iterative time also improved the accuracy of the prediction. The proposed method is further illustrated with its application to the underground cavern of Lvchunba railway tunnel. Based on the surrounding rock’s parameters obtained by back analysis, the displacement of the surrounding rock was predicted. The results showed that the error between numerical calculation value and actual monitoring value was 13.2%,-8.3%,-8.9%,9.4% respectively.

The excavation of foundation pit will change the initial stress state in surrounding soils, which induce the superimposed stress and uneven settlement to the adjacent existing tunnels. Based on the pit excavation engineering in Shenyang, which sits atop the existing tunnels, the risk factors and failure type of shield segment lining were studied through risk analysis theory. According to the uncertainty and spatial variability of soil parameters, the risk accident of existing tunnel was calculated by means of stochastic finite element which is combined Monte Carlo simulation with FEM. The risk of existing tunnel during the pit excavation stage was evaluated respectively under multi-failure patterns. The framework is used to estimate and minimize risks at pit excavation engineering in Shenyang.

The non-pillar sublevel caving method is used in Iron Mine in Banshi. In the mining area, there are many folds and faults, the inclination of ore body changes greatly, and ore and rock are fragmentized. The tunnel often collapsed and the surrounding rock deformation was getting large during the construction stage. Using the data of tunnel surrounding rock deformation, we adopt the neural network method to set up the mapping relation between the tunnel surrounding rock deformation and the project factors, including tunnel deepness, tunnel dimension, measuring time and surrounding rock quality. The analyzing results show that the maximum error between the forecast and the testing data is 13%, which indicates that this method is useful and feasible to the mining engineering. Key words: rock pressure; measure, deformation of the tunnel surrounding rock; neural network; data normalization; mapping