The underground coal mine monitoring system has high costs of large-scale application, the network is not easy to expand, and it is difficult to achieve complete coverage of coal mines. To solve these problems, this paper presents a new data acquisition and transmission method based on 'wired + wireless', 'A, B, C and D class mobile node'. The method, combining the existing wired network and wireless network in the coal mine, using sparse heterogeneous converged networks centred on opportunistic networks, analyses the composition, function, connection law and historical information of four types of node and the system structure. Experimental results show that this method can further realise the dynamic planning of the node interaction path. Then, it predicts and corrects the data of artificial mobile nodes and fixed interaction information in the next cycle. Finally, it achieves a powerful guarantee for internet of things, interaction of objects, intelligent perception, and intelligent processing in the mine safety management.