基于电气距离矩阵特征根分析的主动配电网电压控制分区方法

Voltage Control Partitioning Method for Active Distribution Network Based on Electrical Distance Matrix Eigenvalue Analysis

针对现有电压控制分区方法在确定分区数方面的不足,提出一种基于电气距离矩阵特征根分析的分区方法,并将其应用到主动配电网的电压控制分区中,该方法能通过分析电气距离矩阵的特征根确定最佳分区数。首先,使用牛顿-拉夫逊法进行潮流计算,得到电压/无功灵敏度矩阵,对其进行处理得到增广的电压/无功灵敏度矩阵;然后根据增广电压/无功灵敏度矩阵定义电气距离,得到电气距离矩阵;接着建立描述电气距离矩阵特征根序列的一个指标作为判断依据来确定较大特征根个数,该个数即为分区数;最后,采用基于Ward距离的凝聚层次聚类算法对电气距离矩阵进行聚类,得出分区结果。通过算例分析,验证了所提方法的有效性和合理性。

In view of the shortcomings of the existing voltage control partitioning methods in determining the partitioning number, this paper proposes a partitioning method based on the analysis of electrical distance matrix eigenvalues. The method is applied to the active distribution network （ADN） partitioning, which can obtain the optimal partitioning number by analyzing electrical distance matrix eigenvaiues. Firstly, we use the Newton-Raphson method to calculate power flow, obtain the voltage/reactive power sensitivity matrix, and process it to obtain the augmented voltage/reactive power sensitivity matrix. Then, we define the electrical distance according to the augmented voltage/reactive power sensitivity matrix, and obtain the electrical distance matrix. Besides, we establish an index describing the eigenvalues sequence of the electrical distance matrix as the basis to determine the number of the larger eigenvalues, which is the partitioning number. At last, we use agglomerate hierarchical clustering algorithm based on Ward distance to cluster the electrical distance matrix and obtain the partitioning results. Case simulation is given to prove the effectiveness and reasonability of the proposed method.