基于复杂网络理论的最新进展,提出了一种电力系统无功分区的新算法。该算法由“分裂”和“凝聚”2部算法组合而成。前者利用灵敏度矩阵对电网进行预分裂,确定分区的基本结构。该算法可保证无功分区内部发电机的控制能力,并且显著减少分区合并时的迭代次数。后者基于模块度的概念,构建了分区合并新指标,并据此进行分区合并。该算法可以保证各分区内部的无功就地平衡,同时准确地评估无功分区的划分质量,确定最优分区数目。最后,将提出的算法分别应用于IEEE 39节点系统、IEEE 118节点系统以及上海电网,仿真分析证明了所提出算法的有效性。
Based on the latest advance of complex network theory, the authors propose a novel algorithm for network partitioning, which is composed of two parts, i.e., division and agglomeration. By use of sensitivity matrix, the former pre-divides the network and confirms the primary structure of the areas; the so-called division can ensure the control ability for generators within the network partition and evidently decrease the iteration times while the partitions are merged. Based on the concept of modularity, the latter constructs a new index for partition merging, and the partitions are merged accordingly; the so-called agglomeration ensures that the reactive power within the partition can be balanced in place, the quality of network partitioning can be accurately evaluated and the best partitioning scheme can be determined. Applying the proposed algorithm to IEEE 39-bus system, IEEE 118-bus system and Shanghai Power Grid, simulation results show that the proposed algorithm is effective.