Application of Integrated U-Net Network-Based Microseismic Monitoring and Precision Localization Technology in Cross-Boundary Coal Mining

Cross-boundary mining in coal mines refers to the unauthorized extraction of mineral resources beyond approved mining areas, posing significant threats to coal mine safety and the rational use of resources. Traditional supervision methods are limited in effectively detecting and regulating such concealed activities in a timely manner. To address this issue, this study implemented a pilot application of microseismic monitoring technology at the Hengshan Weiqiang Coal Mine of Shaanxi Yanchang Petroleum Group. By establishing a high-sensitivity, low-noise 4G wireless remote real-time microseismic monitoring system, the study achieved 24/7 monitoring and accurate early warning of mining activities. Ten real-time seismic monitoring systems were deployed, utilizing three-component seismometers and data recording units to continuously monitor the mining area. The collected data was transmitted in real-time to a backend server, where it was processed and analyzed using AI algorithms and geophysical methods. The optimized U-Net neural network was used for P-wave picking, and a grid-point global search algorithm was applied for seismic source location, achieving a location error of less than 5 meters, significantly outperforming traditional methods. The results demonstrated the effectiveness of this system in identifying and locating cross-boundary mining activities, providing a scientific basis for coal mine safety management and resource regulation, and laying the foundation for the broader application of this technology in the future.