摘要
为提升电力系统运行设备数据集成处理效果,研究基于数据驱动模型的电力大数据关联规则挖掘方法。先应用量级差别理论,假定数据的频繁项目集合,通过最小阈值确定其支持度与置信度对电力大数据预处理。然后基于数据驱动模型建立关联矩阵,以非线性自适应控制方式,在极小化处理下聚类分组电力大数据中心。最后通过历史数据中获取信息估计量,动态补偿挖掘电力数据关联规则,实现方法设计。以某省电力系统的营销大数据为测试对象开展实验测试,应用设计方法能够从10 000组数据中挖掘出峰电量、平电量、谷电量三种数据属性,具有较好的应用效果。
In order to improve the data integration processing effect of power system operation equipment,the method of mining association rules of power big data based on data driven model is studied.First,it applies the theory of magnitude difference,assumes the frequent item set of data,determines its support and confidence through the minimum threshold,and preprocesses the power big data.Then,based on the data driven model,the incidence matrix is established,and the power big data centers are clustered and grouped in a nonlinear adaptive control mode under the minimization processing.Finally,the information estimation is obtained from the historical data,and the association rules of electric power data are mined by dynamic compensation to achieve the method design.Taking the marketing big data of a provincial power system as the test object,the application design method can mine the three data attributes of peak electricity,average electricity and valley electricity from 10000 groups of data,with good application effect.
作者
唐远富
陈远扬
TANG Yuan-fu;CHEN Yuan-yang(State Grid Hunan Electric Power Corporation Limited Research Institute,Changsha 410000 China;State Grid Hunan Electric Power Corporation Limited,Changsha 410000 China)
出处
《自动化技术与应用》
2024年第12期110-113,162,共5页
Techniques of Automation and Applications
基金
国网湖南省电力项目(SGHN0000HWJS1900586)。
关键词
数据驱动模型
电力大数据
关联规则
挖掘方法
data driven model
power big data
association rules
excavation method
作者简介
唐远富(1984-),男,博士,高级工程师,研究方向:数据资产运营管理。
