In view of the importance of the active distribution networks in the development of smart grids and the complexity of operation mode and power flows, vulnerability analysis of active distribution network is conducted. The active distribution network is regarded as a directed network, and the direction of the system power flow is taken as the direction vector of the adjacency matrix. Thus, the network betweenness is obtained to measure the role of each node in the network. Given the importance of each node in power transmission, the generator-load factor is considered synthetically, and power factor indicator is proposed. Using the entropy method to determine the weight of each factor, the concept of integrated power betweenness is proposed to identify the vulnerable links of the active distribution network. Based on the time scale of the year, the study of the trend of the vulnerability of the network reveals that the vulnerability of the active distribution network has strong seasonality. In IEEE 33-bus system with distributed power supply and battery energy storage system, the load-loss rate of the network when it is attacked by the vulnerability identification result is calculated, and the correctness of the proposed method is verified. This work is supported by Distinguished Youth Fund of Natural Science Foundation of Hunan Province (No. 2017JJ1011).