Some key nodes in the power system may play a role in the expansion of faults when a largescale interlock failure occurs in the system. In order to improve the speed and accuracy of key node identification, this paper proposes a key node identification method based on PSNodeRank algorithm by improving the PageRank algorithm proposed by Google Company. This method selects the important evaluation index of the key nodes of the power grid, and establishes the directional weighted network model of the power system. Considering the network link direction and the characteristic for the weight of power system network, the PSNodeRank value is proposed to assess the importance of each node. And then the power system partitioning characteristics is utilized to improve the complicated calculation process for the importance of large power grid nodes. The speed of operation is greatly improved and the storage capacity required for the operation is also reduced. Finally, an IEEE 39node system is simulated for verification. It is shown that the proposed method can effectively and accurately identify the key nodes in the power grid and judge their roles in the critical evolution of ACDC power network. This method has a great significance to the critical state evolution of the system.