Abstract: In response to the problems of low equipment collaboration efficiency, low data integration, and incomplete safety monitoring coverage in current coal mine production scheduling and safety management, this paper constructs a digital twin driven ground integrated decision-making and scheduling system for fully mechanized mining face. The system adopts a three-level centralized control system consisting of "intelligent terminals on the working face, control nodes along the channel, and ground control center". It integrates industrial Ethernet and wireless Mesh networks to build a highly reliable communication infrastructure, providing support for data exchange throughout the entire process. In terms of modeling methods, a digital twin of fully mechanized mining equipment was constructed by combining precise 3D models with kinematic based mechanism modeling, thereby achieving dynamic synchronization and precise mapping between physical equipment and virtual systems. Design a multi-source data collection and processing mechanism, develop a collaborative control framework and multi system integration strategy, and ultimately achieve closed-loop intelligent management. The actual application results show that the system can achieve a continuous automatic tracking rate of over 98% for the coal mining machine, a success rate of over 85% for automatic coal discharge, a manual intervention ratio of less than 10%, and a reduction in the number of single shift operators from 5 to 1. In addition, the response latency of the digital twin model is less than 200 milliseconds, the system interface runs smoothly, and the frame rate remains above 30FPS. This research not only expands the application path of digital twin technology in coal mine scheduling scenarios, but also provides an integrated solution with engineering feasibility for promoting the intelligent construction of coal mines.