Abstract: Aiming at the major technical challenges such as complex mine pressure behavior, difficult coordinated control of equipment group and high requirements for mining and transportation balance caused by scale effect in super-long working face of medium-thick and thin coal seam, in order to realize safe and efficient intelligent mining of ' one person, one side and ten million tons ', the construction and application of full-chain intelligent mining system are carried out based on the 450 m working face of Zhuanlongwan Coal Mine of Yankuang Energy (Ordos). Based on the evolution law of strong mining pressure, the ' two-factor ' support control theory of hydraulic support based on dynamic and static load fusion and the group cooperative straightening method based on the ' most lagging support ' benchmark are established. The support cyclic pressure prediction model based on SARIMA algorithm and the high-precision MEMS inertial navigation attitude sensing technology are developed. The dynamic sensing accuracy of support attitude is 0.1 ° RMS, which significantly improves the whole line straightening accuracy of 450 m working face. In the aspect of coal flow control, an intelligent identification system of coal flow load based on UPSNet and YOLO edge computing is constructed, and the Petri net theory is introduced to establish the adaptive control model of production and operation balance, which realizes the intelligent linkage between the traction speed of shearer and the frequency of scraper. Aiming at the precise cutting of shearer, a multi-source fusion navigation digital chassis integrating panoramic laser SLAM, inertial navigation and absolute coordinate target is innovatively constructed, which realizes the absolute coordinate positioning error of shearer ≤ 15 cm, and develops process-driven engine and machine-frame multi-dimensional active anti-collision technology. In addition, a multi-physics coupling digital twin model of hydraulic system and an intelligent pressure stabilization strategy based on IPSO-FNN-PID algorithm are established. The pressure fluctuation of main pipeline is controlled within 1.0 MPa, and the fault predictive maintenance based on digital twin is realized. In order to integrate multi-system collaboration, a “cloud-edge-end” three-level management and control architecture based on 10 Gigabit industrial Ethernet is constructed, which realizes the efficient closed-loop of ground planning, edge control and end-side execution. The industrial application results show that the system makes the average traction speed of the shearer increase to more than 16 m/min, the automatic following rate of the hydraulic support ≥ 97%, the normal operation personnel of the working face decrease by 83%, the straightening accuracy of the whole line increases by 3 times, the recovery rate of the working face increases by about 5%, and the tunneling rate of the ten thousand tons of the panel decreases by 34.3%, which effectively supports the safe, efficient and intensive production of the super-long working face. The research has formed a set of intelligent mining system solutions for medium-thick and thin coal seam super-long working face covering transparent perception, intelligent decision-making, precise execution and collaborative operation and maintenance, which provides a replicable technical path and engineering paradigm for the intelligent construction of mines with similar conditions.