Intelligent analysis and data analysis of microseismic monitoring of tunnel surrounding rock based on optical fiber MEMS vibration sensor

Rockburst hazards in deep underground tunnels pose significant safety challenges to personnel and infrastructure, necessitating reliable monitoring and early warning systems. A time-dependent Gaussian probability density function was developed to analyze spatiotemporal evolution of microseismic events, implemented on a physical model of a trapezoidal roadway in an inclined coal seam (23° dip angle) monitored by 19 MEMS fiber-optic sensors. Acceleration-time analysis revealed distinct phase-dependent characteristics throughout the progressive failure process, from elastic behavior through plastic deformation to eventual failure. Time-space clustering analysis demonstrated pronounced spatial concentration of events, with failure phase events forming a defined cluster 4-6m above the tunnel centerline. Quantitative parameter evolution showed the σ ₍ᵧ₎/ μ ₍ᵧ₎ ratio approaching a critical threshold of approximately 0.3 immediately preceding macro-failure. This methodology enables quantitative delineation of high-risk zones through probability density contours, providing a scientific basis for real-time stability assessment, early warning, and optimized support design in high-stress underground environments.