Tansformer-Customer Relationship and Phase Identification Based on Voltage Sequence Similarity

With the renovation and upgrading of the low-voltage distribution network, the transformer-customer relationship changes frequently. In order to solve the user's hanging wrong phenomenon that happens from time to time in the substation area, an identification method of the transformer-customer relationship and the phase difference in the substation area was proposed by using an improved density-based point sorting recognition clustering structure (abbr. OPTICS). Firstly, the correlation of the distribution network voltage series was analyzed qualitatively, proposing the use of voltage time series as the data basis for analysis and recognition. Secondly, the improved adaptive piecewise aggregate approximation (abbr. APAA) was used to reduce the dimensionality of the voltage series, extracting the low-dimensional vectors that can reflect the voltage characteristics. Thirdly, the extracted feature vectors were clustered by using the improved OPTICS algorithm to identify the transformer-customer relationship and phase relationships of the substation area. finally, an example analysis was carried out based on the actual data of the substation area, which verifies the accuracy of the proposed method.