基于数据挖掘的电力设备状态诊断系统建模
2016-01-20朱付保,霍晓齐,徐显景
基于数据挖掘的电力设备状态诊断系统建模
朱付保, 霍晓齐, 徐显景
(郑州轻工业学院 计算机与通信工程学院, 郑州 450001)
摘要:以电力公司的业务需求为背景,为实现电力设备的状态诊断,以数据挖掘技术中的粗糙集和决策树算法为依据,采用粗糙集和决策树相结合的数据处理模型对电力设备的各属性数据进行了分析处理。综合运用粗糙集和决策树两种数据挖掘算法,通过粗糙集技术进行属性约简,并运用决策树的ID3算法对约简后的数据进行分枝、减枝得到规则集,实现对电力设备工作状态的快速、高效诊断,并根据其工作状态提供决策支持。
关键词:粗糙集;ID3算法;知识库;模型库
中图分类号:TP391
文献标志码:A
DOI:10.3969/j.issn.1671-6906.2015.03.020
Abstract:The system uses the business needs of a power company as the background. For the realization of the state testing of electrical equipment, rough sets and decision tree algorithm belonging to data mining technology is used as the basis. The paper proposes a data processing model combining rough sets with decision tree to realize for power each attribute the data analysis and processing equipment for power each attribute of electrical equipment. Rough sets and decision tree are used in this paper synthetically. Through rough set technology attribute reduction can be realized in this paper. Data which has been reduced is branched and cut branches by the use of the ID3 decision tree algorithm, and then rules in order can be got. At last, the working status of electrical equipment can be diagnosed rapidly and efficiently and the decision support is provided according to their work status.
收稿日期:2014-06-11
基金项目:国家自然科学基金项目(41171341)
作者简介:何渊淘(1983-), 男,河南郑州人,硕士。
文章编号:1671-6906(2015)03-0090-05
随着社会进步和经济的不断发展,工业在人们的日常生活、生产中起到的作用越来越大。迫于各行业对电力能源的巨大需求,电力系统的规模越来越大,并且不断朝着超高压、跨地区的趋势发展。与此同时,当变压器、断路器等电力设备出现故障时,给人们的生活和工业生产带来的影响也日益增大,这必然对电力设备安全、稳定、可靠的运行提出越来越高的要求。
然而,电力行业传统的状态检测效率并不高。计划检修作为常规的检测方法机械地按固定周期对所有电力设备进行检测诊断,工作十分繁琐,且存在工作量大、维修滞后、检修过剩等问题。……
