The multi-disciplinary cross-integration of Internet of things (IoT) technology provides a new way to thoroughly solve the problem of mine safety production. The real-time accurate perception and intelligent analysis of mine “human-machine-environment” status information is the basis of mine IoT application. The technological architecture of the IoT was analyzed for mine safety production, the virtual-real fusion perception theory model based on CPS and DT was established to solve key issues in mine IoT, such as perception theory, intelligent sensing technology, standardization method and large data analysis service. A unified description method for mine “human-machine-environment” information was put forward, the multi-parameter sensors and intelligent autonomous sensing devices were developed, and the mine IoT coding and interaction protocol standards were standardized. In addition, it was designed for the mine IoT converged communication network service architecture, the application service system of high-efficiency access and deep analysis of large data in mine IoT was constructed, which realized the comprehensive and accurate perception, real-time and efficient interaction and intelligent analysis application of mine safety production status information. It also provides strong scientific and technological support for the mine safety prevention preparation, monitoring and early warning, situational prejudgment, comprehensive protection and other technical levels.