Abstract: Coal is the key energy source in China, and it is of great strategic significance to ensure the safe and efficient mining of coal resources in China. China’s coal resources are mainly mined by shaft mining, the production environment is harsh and the labour intensity of the workers is high, and intelligent mining has become a necessary path for the high-quality development of China’s coal industry due to its conceptual connotation of reducing manpower and improving efficiency. Hydraulic support is the key support equipment for underground mining, and its support performance is affected by the support posture of the body and the associated posture of the group at the same time. Under the influence of complex mining conditions, it is difficult to obtain the support attitude information of hydraulic support, so how to intelligently and accurately perceive the spatial support attitude of hydraulic support group is one of the major problems restricting the intelligent construction. In order to accurately describe the spatial support attitude of the hydraulic support group in the working face, firstly, the spatial connection relationship of the hydraulic support group is analysed, and then the hydraulic support self-posture perception model based on the multi-rod double-drive and the frame group posture perception model based on the spatial multi-line cooperative posturing are established, and then the influence of different solution strategies on the performance of the self-posture model is analysed. Subsequently, the digital twin environment of hydraulic support group in the working face was constructed with the help of 3D interactive platform Unity3D, and a data-driven visualisation system based on the real scene in the working face was developed. Finally, an engineering validation of the proposed method was carried out in a mine in Shandong Province, and the results show that the proposed spatial support attitude twin reconstruction method can accurately map the real spatial support attitude of the actual hydraulic support group, which solves the problem that the support attitude information of the hydraulic support body is easily interfered by the vibration in the field and the associated attitude information of the support group is missing. The proposed method provides ideas for the virtual reproduction of the whole scene of the operation behaviour of multi-intelligent bodies in the working face and the intelligent construction of the working face.