Gas insulated switchgear (GIS) equipment has compact structure, closed space and highly integrated functions. Limited by the complexed structure, it is difficult to quickly locate the fault area and take maintenance measures. Hence, the modular division is necessary for the secure and stable operation and reliability assessment. In the paper, the Design Structure Matrix (DSM) model is established with the comprehensively analysis of the correlation between different functional elements of GIS equipment. By constructing the fitness function and using the genetic algorithm for clustering operation, the optimal division result of GIS is obtained. The results show that the DSM model based on GIS functional elements clusters significantly, and the fitness function converges to the optimal value 0.25. Compared with traditional division methods, the proposed method in this paper has the advantage of multi-view feature association and stable clustering effect.