摘要
在飞速的推进和发展智能电网的过程中,各种监测的数据数量不断的变多,采用传统的电力数据的处理系统已经无法消化处理这些数据。文章结合大力大数据技术的现状分析,从时空的内存占用、时间运行速率2个维度探究其存在的技术短板,并针对K-means算法对于初始聚类中心过分依赖性问题,设计一种多初始聚类中心、多机组并行处理的改K-means算法,解决了传统K-means算法的弊端性,且基于算例实验验证了改进K-means聚类算法显著提升运行效率、聚类有效性,可优化电力大数据分析平台的应用性能。
In the process of rapid advancement and development of smart grids,the amount of various monitored data is constantly increasing,and traditional power data processing systems have been unable to digest and process these data.Based on the analysis of the current situation of big data technology,this paper explores the technical shortcomings of big data technology from two dimensions of space-time memory occupation and time running rate.Aiming at the problem of over dependence of K-means algorithm on the initial clustering center,this paper designs a modified k-means algorithm with multi initial clustering center and multi unit parallel processing,which solves the disadvantages of traditional K-means algorithm,and is based on the example It is verified that the improved k-means clustering algorithm can significantly improve the operation efficiency and clustering effectiveness,and optimize the application performance of power big data analysis platform.
作者
施雯
Shi Wen(Foshan Power Supply Bureau of Guangdong Power Grid,Foshan 528000,China)
出处
《现代科学仪器》
2019年第5期92-94,共3页
Modern Scientific Instruments
