The invention discloses a low-voltage transformer area user topology identification method based on a t-SNE dimension reduction technology and BIRCH clustering. Topological information of a low-voltage transformer area has important guiding significance for transformer-user relationship error correction and line loss governance analysis, however, current error correction of topological archives mainly depends on on-site troubleshooting of electrical employees, manpower and material resource consumption is large, and troubleshooting efficiency is low. In the background, the invention provides atopology identification method based on voltage measurement data of an intelligent electric meter. The method comprises the following steps: firstly, performing dimension reduction processing on original load data by adopting a t-SNE technology, and solving the redundancy problem caused by too high original load feature dimension of a transformer area user; and then, identifying the phase to which the single-phase user belongs and the secondary access meter box in the transformer area by using a BIRCH method. The result of example analysis shows that the method provided by the invention has ahigher recognition rate for the topology of the transformer area, and can provide reference for the topology recognition and troubleshooting of the low-voltage transformer area.