There is a shortage of traditional methods for extracting the typical user load curve and the feature extraction of the load curve in the current user load curve analysis method. This paper proposes a user load curve analysis method based on time series data mining. First, the method reduces the dimension of the load curve of a single user based on piecewise aggregate approximation, and re-expresses the load curve of the user for a period based on the symbolic aggregate approximation method, and represents the user load curve with a symbolized sequence to extract the typical load curve of the user. Then, combined with the load characteristics and time series characteristics of the typical load curve of different users, based on the k-means algorithm, the typical load curves of different users are clustered to analyze the power consumption characteristics of different types of users. A UCI test data set is used for case analysis, and the results show that the proposed method can mine the typical power consumption characteristics of users and improve the user load curve analysis efficiency and clustering quality. This work is supported by the Science and Technology Project of China Southern Power Grid Co., Ltd. (No. GDKJXM20173251).