Abstract: Aiming at the problems of relatively high energy consumption, lagging monitoring, insufficient response to safety incidents and others in traditional mines, a key technology system for green intelligent mine is constructed. Relying on multi-source perception, edge computing and 5G communication technology, a three-layer collaborative structure of "perception-control-supervision" is formed, and a high-precision sensor network, adaptive control algorithm and safety environment integrated supervision system are designed; Through multi-protocol fusion and distributed data stream processing, real time acquisition and dynamic optimization of production parameters are achieved. Taking the operational level before the transformation of Zhaiyadi Coal Mine in Liulin County as the comparison benchmark, the application verification indicates that the unit energy consumption is decreased from 1.20 kW·h/t to 0.97 kW·h/t (with a decrease of 19.6%), and the carbon emission intensity is decreased from 3.5 kgCO₂/t to 2.73 kgCO₂/t (with a decrease of 22%); The response time for safety incidents is shortened from 6.0 minutes to 1.8 minutes (with a reduction of approximately 70%). This system is designed for special working conditions in high gas mines, and can achieve deep integration of green and intelligent construction, which provides a technical support for the systematic promotion of intelligent mines in the coal industry.