Abstract: The situation awareness of smart distribution network is an important foundation for the reliable, economic and safe operation of the distribution system. The scale and type of its data are growing rapidly, showing typical characteristics of power big data. As the data collected and processed by the micro PMU system of smart distribution network has a tendency of massive growth, this paper introduces a situation awareness method of smart distribution network based on micro PMU data mining, which can judge the system security state quickly and accurately. The method firstly uses moving and dynamic time windows to further mine the typical characteristics of various events, and then, performs hierarchical labeling according to region selection and event types, reducing the computational dimension of machine learning algorithms and improving computational efficiency. Afterwards, this paper designs three types of classifiers and compares them with the other two types of classifiers. Through the test data of the example, it is concluded that the performance indicators of the proposed classifier are excellent, which shows that the classifier has important reference value for the establishment of situation awareness system of smart distribution network.