Topology Identification Method for Low-Voltage Stations Based on Mixed Integer Programming

Accurate and reliable station topology relationships are a key factor ensuring that low-voltage (LV) distribution networks operate stably and reliably. A large number of affiliation errors exist in the customer-transformer relationships of LV stations. In this paper, firstly, combining the data of electricity information acquisition system and the basic archive data of marketing system, we calculate the highly negative correlation of line loss and the obvious existence of abnormal customer-transformer affiliation of the station area. After that, binary variables are introduced to accurately represent the affiliation relationship of the abnormal station area, and a mixed integer model is constructed with line loss sum of squares minimization as the optimization objective for the station area customer-transformer calibration, and the model is solved by the branch-and-bound method. Then the topology relationship with the highest probability of decreasing the highest priority is obtained by using the integer partition method. Finally, the reliability evaluation index is developed to evaluate the reliability of the station topology identification results.