摘要
针对电网网损数据样本聚合精度较低、计算精度不高的问题,提出一种基于数据挖掘计算的电网网损评价算法。该算法基于最近聚类算法原理,通过聚类中心优化调整、最远数据点二次聚类、阈值动态选取等技术,使得基于电网断面采集数据的聚类质量得到有效提升,并使基于该聚类结果的电网网损评价算法的精度得到可靠保证。仿真测试证明了,通过该算法能得出拟合度较高的电网网损曲线。
Aiming at the problems of low polymerization accuracy and low calculation accuracy of power loss data samples,this paper proposes a power loss evaluation algorithm based on data mining.On the basis of the nearest neighbor clustering algorithm,through optimization and adjustion of the clustering center and secondary clustering of the farthest data points based on dynamic selection technology of clustering threshold,the clustering quality of the collected data based on power grid tie-line is effectively improved,and the accuracy of power loss evaluation algorithm based on the results is reliable guaranteed.The simulation test proves that,with this algorithm,the power loss curve with a high fitting degree can be obtained.
作者
陈忠
吴靓
CHEN Zhong;WU Jing(Guangdong Polyttechnic of Water Resources and Electric Engineering,Guangzhou 510635,China)
出处
《电工技术》
2018年第9期3-6,共4页
Electric Engineering
关键词
数据挖掘
聚类算法
网损
聚类中心
聚类阈值
data mining
clustering algorithm
power loss
clustering center
clustering threshold
