A Constraint-Based Clustering Method for Automatic Voltage Control Soft Partitioning and Partition Result Evaluation

摘要

In order to better meet the requirements of automatic voltage control, this study proposes a constraint-based clustering method for power grid soft partitioning. By pre-setting boundary conditions into the clustering algorithm, this method solves the problems that exist in the traditional method, in which one needs to artificially determine the number of partitions and a whole partition contains isolated bus only. Next, the study proposes a method to further improve the partition result by regulating the few unreasonable buses. Then, this study proposes a partition result evaluating method based on electrical gravitation. This evaluating method provides a reliable basis for verifying whether the boundary conditions are reasonable and distinguishing the merits of different partitioning methods. At last, the partition result of this study method is compared with the existing method. The comparison results show that this study method can better meet the requirement of automatic voltage control.