摘要
为了解决目前电力调控数据中心机房设备众多且日趋复杂而难以有效管理的问题,提出了一种改进的快速密度峰值聚类算法,通过对机房设备产生的海量运行数据进行深度分析和挖掘,从而完成精确故障定位的目的。该算法在原始快速密度峰值聚类算法(FDPC)的基础上,采用带有递减的本地自适应窗口(LAW)的迭代覆盖图(OM)程序来构建多维密度图和决策图。通过将构建的密度图与FDPC进行结合显著减少了后续步骤处理的复杂度。此外,提出的算法对参数变化不敏感,因此,可以对聚类簇中心的选择过程进行持续优化。实验结果表明,提出的算法对电力调控数据中心机房设备的故障检测达到了较高的准确率。
To solve the problem of the increasing complexity and difficulty in effective management of the numerous equipment in the power dispatching data center computer rooms,an improved fast density peak clustering algorithm was proposed to perform deep analysis and mining of the massive operational data generated by the equipment in the computer room to accomplish the purpose of accurate fault location.Specifically,on the basis of the original fast density peak clustering algorithm(FDPC),an iterative overlay(OM)program with a decreasing local adaptive window(LAW)was used to build and establish a positioning map and a multi-dimensional density map.By combining the constructed density map with FDPC,the complexity of subsequent processing is significantly reduced.In addition,the proposed algorithm is insensitive to parameter changes,therefore,the selection process of cluster centers can be continuously optimized.The experimental results show that the proposed algorithm achieves a high accuracy rate for the fault detection of the equipment in the power dispatching data center computer room.
作者
王婷
王其兵
燕争上
张秀丽
徐利美
石新聪
朱燕芳
强彦
郝伟
WANG Ting;WANG Qibing;YAN Zhengshang;ZHANG Xiuli;XU Limei;SHI Xincong;ZHU Yanfang;QIANG Yan;HAO Wei(State Grid Shanxi Electric Power Research Institute, Taiyuan 030000, China;State Grid Shanxi Electric Power Company, Taiyuan 030000, China;College of Information and Computer, Taiyuan University of Technology, Jinzhong 030600, China)
出处
《太原理工大学学报》
CAS
北大核心
2020年第6期860-866,共7页
Journal of Taiyuan University of Technology
基金
山西省自然科学基金面上项目(201801D121139)。
关键词
电力调控数据中心
故障定位
密度峰值聚类
机房设备
异常值检测
power control data center
fault location
density peak clustering
computer-room equipment
outliers detection
