一种基于数据挖掘的无人飞行器故障分析方法  

A Fault Analysis Method Based on Data Mining for Unmanned Aerial Vehicle

作　　者：孟魁杰[1] 董莹[2] 赵宗涛[2] 

机构地区：[1]陕西师范大学计算机科学学院,陕西西安710062 [2]第二炮兵工程学院,陕西西安710025

出　　处：《计算机技术与发展》2010年第6期225-227,232,共4页Computer Technology and Development

基　　金：中国人民解放军第二炮兵预研项目(2007014016)

摘　　要：无人飞行器的设计与应用包含了高科技技术的诸多方面,是一个极为复杂的机电系统。为保障可靠飞行,而对其进行故障检测是一件必不可少的工作。提出一种基于故障库(FRDB)的数据挖掘技术,对无人飞行器进行了故障分析研究,并设计了相应的算法。从FRDB数据库中采用选择、投影操作,分离出相关部件的超差字段,形成相应的数据文件,以此建立联机统计分析的基础。而后,又对发生故障的分布规律进行统计分析。测试中重点放在无人飞行器控制系统的各部件,为实现统计分析的自动化和半自动化,必须重视测试数据的规范化处理和数据库的建立。Design and application of unmanned aerial vehicle include many aspects of high technology,and it is a highly complex electro-mechanical system.To ensure the reliability of flight,it is an essential work to perform fault detection.A data mining technology based on FRDB is proposed to study fault analysis to unmanned aerial vehicle,and also the corresponding algorithm is designed.Selection and projection operation from FRDB are used to separate the super-bad fields of the relevant part and form the corresponding data files to establish the basis of statistical analysis.Besides,statistic analysis is performed to the distribution of the fault.In testing procedure,the components of the unmanned aerial vehicle control system are paid high attention.In order to achieve the automation and semi-automation of statistical analysis,normalized treatment to the testing data and the database building must be paid attention.

关 键 词：数据挖掘 无人飞行器 故障分析 

分 类 号：TP311[自动化与计算机技术—计算机软件与理论]