Abstract:

To address the problems of early warning systems for various disasters in underground coal mines, including fragmentation, isolated information, and slow response, an architecture for a comprehensive disaster prevention and control platform in underground coal mines was developed. The "1+5+N" business framework and a layered technical system were adopted by the architecture. Five types of disaster monitoring data and several business subsystems such as ventilation, drainage, and personnel positioning were integrated. A base for five intelligent early warning models was established, including trend prediction, disaster simulation, multi-factor risk early warning, decision support, and forecast and early warning. An application validation was also carried out. The results indicate that by using microservices and distributed computing, the platform enables early identification, dynamic analysis, and early warning of risks. The architecture effectively improves data integrity and fusion efficiency, with the forecast success rate increasing by 35% and timeliness improving by 60%. A whole-process flow from risk perception to linkage control is successfully realized.

Key words: underground coal mine, comprehensive disaster prevention and control, platform architecture, intelligent early warning model, linkage