动态粒度结合中心点算法在电力设备缺陷管控中的应用研究  

Application of dynamic granularity combined with center point algorithm in power equipment defect control

作　　者：万少明 代金磊 WAN Shaoming;DAI Jinlei(State Grid Xinjiang Electric Power Co.,Ltd.Hami Power Supply Company,Hami 839000,China)

机构地区：[1]国网新疆电力有限公司哈密供电公司,新疆哈密8839000

出　　处：《中国高新科技》2024年第3期79-81,共3页

摘　　要：针对电力设备缺陷管控问题,文章研究提出了一种基于改进k-中心点聚类算法与动态粒度的电力设备缺陷管控模型。首先,利用改进的k-中心点聚类算法对设备缺陷数据进行聚类处理;然后,将动态粒度与改进算法进行结合,用于构建缺陷管控模型。结果表明,缺陷管控模型的数据聚类正确率为93.07%,聚类效率能够达到90.07%,同时数据识别准确率、召回率和F1值分别为93.27%、93.52%和0.951,均优于对比方法。这说明研究构建的电力设备缺陷管控模型显著可以提高设备的可靠性和稳定性。Aiming at the problem of power equipment defect control,the study proposes a power equipment defect control model based on the improved k-center point clustering algorithm and dynamic granularity.The study first uses the improved k-center point clustering algorithm to cluster the equipment defect data;then combines the dynamic granularity with the improved algorithm for constructing the defect control model.The results show that the data clustering correct rate of the defect control model is 93.07%,and the clustering efficiency can reach 90.07%,meanwhile,the data recognition accuracy,recall rate and F1 value are 93.27%,93.52%and 0.951 respectively,which are better than the comparison method.This indicates that the power equipment defect control model constructed in the study can significantly improve the reliability and stability of the equipment.

关 键 词：动态粒度 k-中心点聚类算法 电力设备 缺陷管控 

分 类 号：TM507[电气工程—电器]