Accurate topology is the basis for good management and safe operation of a distribution network. There are problems in that the existing distribution network topology identification methods cannot determine the direction of the network power flow and the topology change detection method is greatly affected by load mutations. This paper proposes a distribution network topology change detection and identification method based on a directed adjacency matrix. First, a distribution network topology identification model based on a directed adjacency matrix is proposed. The model describes the unknown topology of the distribution network as an uncertain directed adjacency matrix to realize the topology identification of the distribution network and determine the power flow direction. Secondly, a topology change detection method based on a node voltage adjacency matrix is proposed. This method judges whether the topology has changed by analyzing the downward trend of the voltage amplitude of adjacent nodes. In addition, the method proposed can narrow the scope of unknown topology and reduce the variable scale of the topology recognition problem. Finally, simulation analysis verifies the correctness and effectiveness of the proposed method. This work is supported by the National Natural Science Foundation of China (No. 51907114).