基于多值属性Apriori算法的飞机IDG故障分析  

Fault Analysis of Aircraft IDG Based on Multi-valued Apriori Algorithm

作　　者：孔祥芬[1] 张利寒 刘敬赟 王杰 KONG Xiang-fen;ZHANG Li-han;LIU Jing-yun;WANG Jie(College of Aeronautical Engineering, Civil Aviation University of China, Tianjin 300300,China)

机构地区：[1]中国民航大学航空工程学院,天津300300

出　　处：《组合机床与自动化加工技术》2021年第2期69-72,81,共5页Modular Machine Tool & Automatic Manufacturing Technique

基　　金：航空科学基金项目(20170267002);民航机场群智慧运营重点实验室开放基金项目(KLAGIO20180302)。

摘　　要：为了更为高效地挖掘海量航空维修数据中隐含的信息,提升飞机故障分析能力及深度维修能力,通过轮廓系数K-means聚类和多值属性Apriori关联规则算法挖掘飞机IDG部件各故障因素之间的关联规则。首先,采集B737飞机故障维修数据,进行预处理并筛选出与IDG部件相关的故障数据信息;其次,利用K-means聚类和多值属性Apriori关联规则算法对IDG部件故障数据进行聚类分析和关联规则挖掘,得到主要故障现象及各因素之间的强关联规则,和传统Apriori算法对比,多值属性Apriori算法的运行时间更短、效率更高;最后,结合各组件工作机理对B737飞机IDG部件提出了相关预防性维修建议。In order to mining hidden information of massive aviation maintenance data efficiently,improve aircraft fault analysis and deep maintenance capability,Mining the association rule between fault factors of aircraft IDG components by contour coefficient K-means clustering and Apriori association rule algorithm based on multi-valued attributes.Firstly,collecting fault maintenance data of B737 aircraft,then perform pre-processing and filter out fault data information associated with IDG components;secondly,performing cluster analysis and association rule mining for failure data of IDG component using K-means Clustering and Multi-Valued attribute Apriori Algorithm,obtain major failure phenomena and strong association rule between various factors,compare with the Apriori algorithm,Multi-valued attribute Apriori algorithm has shorter runtime and higher efficiency;finally,combined with the working mechanism of components to provide suggestions for preventive maintenance of B737 aircraft IDG components.

关 键 词：数据挖掘 多值属性Apriori算法 故障分析 

分 类 号：TH16[机械工程—机械制造及自动化] TG65[金属学及工艺—金属切削加工及机床]