【摘要】 低压配电网变-相-户关系的自动识别是实现智能低压配电网的基础。针对目前低压配电网拓扑连接关系识别难、用户信息更新换代快、已知的连接关系易过时的问题，给出一种基于先验知识及数据相关性的户变关系识别方法。首先以配电变压器某一相线为例，推导相关公式，分析电压特性及相关先验知识。然后借助电压相关系数，确定初步的户变关系及嫌疑用户集合，再由分类矩阵和阈值系数向量确定近端用户集合。最后在此基础上，剔除嫌疑用户集合中的错误用户，对剩余嫌疑用户采用相关性均值、方差处理，实现户变关系及错误用户正确归类。选取南昌市某地区实际数据展开算例分析，结果证明了本文给出方法的可行性和有效性，能有效解决户变关系问题。更多还原

【Abstract】 The automatic identification of transformer phase household relationship in low-voltage distribution network is the basis of realizing intelligent low-voltage distribution network.In view of the problems that it is difficult to identify the topological connection relationship of low-voltage distribution network, the update of user information is fast, and the known connection relationship is easy to be outdated, a household transformer relationship identification method based on prior knowledge and data correlation was proposed.Firstly, taking a phase line in the station area as an example, the relevant formulas were deduced, and the voltage characteristics and relevant prior knowledge were analyzed.Then, with the help of voltage correlation coefficient, the preliminary household change relationship and suspected user set were determined, and then the near end user set was determined by classification matrix and threshold coefficient vector.Finally, on this basis, the wrong users in the suspected user set were eliminated, and the correlation mean and variance were used for the remaining suspected users to realize the user change relationship and the correct classification of wrong users.The results showed that the method given in this paper was feasible and effective, and can effectively solve the problem of household transformer relationship.更多还原