Hyperspectral imaging provides a novel approach for intelligent geological perception in tunnelling and underground engineering due to its high spectral resolution, nondestructive nature, and combined spectral-spatial information. However, in confined underground spaces, noise is often introduced by short exposure times, low illumination, and dust, and limited spatial resolution can cause mixed pixel effects, complicating data processing. This study presents an underground hyperspectral imaging-based mineral mapping method that achieves wall-rock visualization and semi-quantitative mineral mapping through image denoising and spectral unmixing. A spatial-spectral recurrent transformer U-Net is developed to reduce noise by leveraging spectral band correlations and nonlocal spatial-texture dependencies. A Dirichlet-based mixed pixel simulation is used to address spectral mixing, with the N-FINDR algorithm identifying endmember minerals, and the fully constrained least squares method to estimate mineral abundances. When applied to a water diversion tunnel in Shanxi, the method generates spatial distribution maps of dolomite and calcite. The experimental results confirm its effectiveness for intelligent geological logging and subsurface geological feature analysis.

Leakage disasters in shield tunnels frequently occur, leading to severe consequences such as tunnel collapse, road collapse, and building destruction. Since it is difficult to record the accident evolution process onsite, it is necessary to reproduce it through credible numerical simulations. However, traditional numerical methods face technical bottlenecks when simulating water-sand inrush in shield tunnels due to challenges such as large deformation analysis and fluid-structure coupling, making it difficult to simulate the process of disaster progression. To address this issue, a Coupled Eulerian-Lagrangian (CEL) method incorporating seepage analysis, referred to as the S-CEL method, was proposed to simulate the interaction between water, soil, and a shield tunnel during a disaster. A refined three-dimensional numerical model was developed using the S-CEL method to simulate the water-sand inrush process. The generation sequence of new leakage points at the segment joints and the mechanisms driving the progression of the disaster were revealed. New leakage points were progressively generated along the longitudinal direction of the tunnel. As the number of leakage rings increased, the amount of soil loss increased rapidly. This led to severe uneven settlement and dislocation deformation of the tunnel. A channel steel was introduced to reinforce the tunnel in the numerical simulation to mitigate or decelerate the progression of the leakage disaster. The connection method between the channel steel and tunnel segments was found to be pivotal to the strengthening effect. Employing only bolt anchoring showed limited efficacy, while enhancing the segment-steel interface with epoxy resin achieved much better performance in mitigating disaster progression.