Hierarchical optimization of acoustic emission sensor placement for cross-fault tunnel linings: A physics-constrained and damage-probability-driven approach

Tunnel linings crossing active faults are prone to highly non-uniform damage under the coupled effects of seismic loading and fault dislocation, posing challenges to effective acoustic emission (AE) monitoring. Sensor layouts based on geometric rules or engineering experience may have difficulty simultaneously ensuring adequate coverage of damage-prone regions and effective signal reception. To address this issue, this study proposes a physics-constrained, damage-probability-driven hierarchical optimization framework for AE sensor placement in fault-crossing tunnel linings. Numerical simulations are used to predict the spatial distribution of lining damage and identify priority monitoring regions, while AE wave-propagation analysis is employed to quantify the constraint imposed by signal attenuation on the effective monitoring range. Based on these results, sensor locations are optimized in continuous space using the RIME algorithm. The results show that damage-critical regions are mainly concentrated on both sides of the fault zone, while AE signal attenuation is affected by the tunnel cross-sectional geometry and propagation path. At the representative -17 dB threshold, compared with uniform placement, the optimized layout reduces the sensor count by 16.7%, increases critical-zone coverage from 97.4% to 100%, and improves the minimum received signal quality by 1.96 dB. The results indicate that integrating damage distribution with signal-propagation constraints helps improve monitoring coverage and signal reception in critical regions while reducing the required number of sensors, thereby providing a design basis for the rational allocation of AE monitoring resources in fault-crossing tunnels.