To seek key nodes and improve network robustness, the dynamic convergence complex network model and its node importance evaluation method are proposed for wired and wireless integrating layered networks. Considering characteristic of dynamic convergence complex networks, parameters including edge connection probability, path connection probability, network connection probability, convergence node proportion, convergence node distribution and convergence path proportion are designed. Based on node importance evaluation indexes in single-layer complex networks, the node degree centrality, node betweenness centrality and node convergence centrality in dynamic convergence complex networks are presented. Node convergence centrality of convergence nodes indicates their contribution to network convergence, and that of non-convergence nodes indicates their auxiliary effect to network convergence, especially they are used as relay nodes among convergence nodes. At last, node importance evaluation is implemented considering network topology structure and its dynamic convergence characteristic. Typical example results of improved dynamic convergence kite networks show that the proposed method can comprehensively depict the node importance in dynamic convergence complex networks. Simulation network composed of fiber communication network and satellite communication network is designed by NS2, further indicating the effectiveness of the proposed method.