摘要
大型工业控制系统中存在海量的运维数据,为了降低高维数据最大频繁项集对数据挖掘结果的影响,提升控制系统性能和环境安全,在关联规则的基础上,提出了一种新的数据挖掘方法。在给定的事务数据库中,通过不断改变数值,使得支持度和置信度始终保持最小值,保证关联规则为强关联;经过预处理,将数据转换为离散型数据,运用Apriori算法和DLG算法产生频繁项目集,构建关联图得到频繁项集。当不再产生新的项集时终止计算,所得项集即为最终的数据挖掘结果。在TEP仿真系统上展开实验,结果表明,所提方法可以准确挖掘到控制系统运行过程中的异常数据,并以明显的波动提醒工作人员及时查看。
There are massive operation and maintenance data in large industrial control systems.In order to reduce the impact of the maximum frequent items of high-dimensional data on mining results and improve the performance of control systems and environmental security,a new data mining method is proposed based on association rules.In a given transaction database,by constantly changing the value,the support and confidence are always kept to the minimum,so as to ensure that the association rules are strong association.After preprocessing,the data are transformed into discrete data,the frequent item set is generated by Apriori algorithm and DLG algorithm,and the frequent item set is obtained by constructing association graph.When no new item set is generated,the calculation is terminated,and the resulting item set is the final data mining result.An experiment is carried out on the TEP simulation system.The results show that the proposed method can accurately mine the abnormal data in the operation of the control system,and remind the staff to check it in time with obvious fluctuation.
作者
李军
LI Jun(CNNC Jianzhong Nuclear fuel Co.,Ltd.,Yibin 644000,China)
出处
《微型电脑应用》
2023年第9期167-170,共4页
Microcomputer Applications
作者简介
李军(1975-),男,本科,高级工程师,研究方向为计算机应用。
