摘要
为更好解决二级电压控制分域问题,研究一种基于可视化自组织特征映射网络聚类技术的分域方法。根据自组织特征映射网络的拓扑保持、信息融合和高维数据矢量量化的特性,建立系统电压分域可视拓扑映射模型,以拓扑映射图为基础进行二级电压控制分域。在此基础上,提出以熵作为测度搜索主导节点。该分域算法融合了节点电压、相角和灵敏度等信息,利于分区间的控制解耦。该方法可离线训练,在线计算。
In order to solve the problems of the secondary voltage control partition effectively, an approach to partition based on the visual self - organizing feature mapping network clustering technique was presented. Based on the characteristics, such as topology holding, information bond andhigh- dimension data vector quantization, which the self- organizing feature mapping network has, the visual topology mapping pattern used for partitioning system voltage was set up meanwhile the secondary voltage control partition was carried out basing on the topology mapping image, after that It is presented that entropy is used to search dominant node as measuring variable. The approach to partition combined with node voltage, phase angle and sensitivity etc, in favor of control decoupling between partitions.
出处
《黑龙江电力》
CAS
2006年第3期180-185,共6页
Heilongjiang Electric Power
关键词
控制分区
主导节点
模式聚类
熵测度
U-矩阵
control partition
dominant node
mode - clustering
entropy measure
U - matrix
