摘要
动力监测采集的数据量巨大,蕴藏信息丰富;仅用于越限故障报警,蕴含知识(规则)未能合理挖掘和运用。为此,基于上三角矩阵改进了关联规则频繁项集算法,可精简候选项集,减少数据库遍历和扫描次数,提高规则挖掘效率,并应用于某动力监测数据的关联规则分析,实例验证了改进算法的可行性和有效性,为动力监测数据挖掘、提取关联规则提供了一有效技术途径。
Power monitoring device collects large amounts of data which is only used for trouble alarm. The data contain a wealth of knowledge(rules), but have not been effectively excavated. Therefore, based on the upper triangular matrix, an improved association rule algorithm is introduced. The algorithm can streamline the size of candidate sets and minimize the number of traversing and scanning for the database. So the efficiency of rule mining is significantly improved. The feasibility and validity of the improved algorithm are verified by an experiment which is conducted in a power monitoring data through association rule analysis. It provides an effective technical approach for power monitoring, data mining and extraction of association rules.
引文
[1]孙彤.关联规则在船舶变压器故障诊断中的应用[J].舰船科学技术,2017,39(2A):31-33.
[2]范苏颖.浅析计算机数据挖掘技术[J].信息系统工程,2017(2):94.
[3]艾金勇.图书馆读者借阅行为的关联规则挖掘研究[J].情报探索,2017(1):40-43.
[4]白莹莹,申晨晨.基于关联规则挖掘的Apriori改进算法[J].电子技术与软件工程,2017(3):203-204.
[5]陈志飞,冯钧.一种基于Apriori算法的优化挖掘算法[J].计算机与现代化,2016(9):1-2.
[6]黄龙军,段隆振,章志明.一种基于上三角项集矩阵的频繁项集挖掘算法[J].计算机应用研究,2006(11):25-26.
[7]边根庆,王月.一种基于矩阵和权重改进的Apriori算法[J].微电子学与计算机,2017,34(1):136-139.
[8]邹永久,张跃文,孙培廷,等.基于RKGM-AR模型的船舶主机排气温度预测[J].船舶工程,2015,37(7):46-49.