欢迎来到天天文库
浏览记录
ID:52102449
大小:1.87 MB
页数:8页
时间:2020-03-22
《基于Markov特征的油气管道泄漏检测与定位方法.pdf》由会员上传分享,免费在线阅读,更多相关内容在行业资料-天天文库。
1、第38卷第4期仪器仪表学报Vol38No42017年4月ChineseJournalofScientificInstrumentApr.2017基于Markov特征的油气管道泄漏检测与定位方法刘金海,臧东,汪刚(东北大学信息科学与工程学院沈阳110004)摘要:针对传统的基于压力信号的管道泄漏检测方法误报率和漏报率偏高,同时定位误差较大的缺点,设计了一种基于Markov特征的管道泄漏检测与定位方法。首先,将管道压力数据构造为Markov链的形式,并提取其动态特征;然后,将所提取的特征应用于NeymanPearson异常检测方法之中,检测全部压力数据样本的状态,并对检测
2、到的异常样本进行同源信号匹配,修正检测结果;最后,将相似性定位方法与连续小波定位方法结合,确定管道首末两端响应压力变化的时间差,并根据管道长度和压力波传输速度等信息,对泄漏源定位。所提方法能应用于小泄漏和缓慢泄漏的检测与定位,易于实现,误报率与漏报率显著降低,定位精度提高。通过对历史数据的分析,验证了所提方法的可行性和有效性。关键词:油气管道;Markov链;NeymanPearson异常检测;相似性定位;连续小波定位中图分类号:TH878TH865TE832TE88文献标识码:A国家标准学科分类代码:440.55Leakagedetectionandlocationmet
3、hodofoilandgaspipelinesbasedonMarkovfeaturesLiuJinhai,ZangDong,WangGang(SchoolofInformationScienceandEngineering,NortheasternUniversity,Shenyang110004,China)Abstract:Aimingatshortcomingsoftraditionalleakagedetectionmethodsbasedonpressuresignal,suchashighfalsepositiverateandfalsenegativerate
4、,andlargelocationerror,aleakagedetectionandlocationmethodofoilandgaspipelinebasedonMarkovfeatureisdesigned.Firstofall,aMarkovchainisbuiltfromthepipelinepressuredata,atthesametime,thedynamicfeaturesareextracted;then,inordertodetectallpressuredatasamplestates,theNeymanPearsonanomalydetection
5、methodisappliedtothedynamicfeatures;meanwhile,thedetectionresultsarecorrectedthroughcheckingwhetherthedetectedabnormalsamplesarematchingwiththesamesourcesignal;last,similaritylocationmethodandcontinuouswaveletlocationmethodarecombinedtodeterminethetimedifferenceofthepressurechangesatbothend
6、softhepipe,andtheleakageislocatedaccordingtheinformationofpipelinelengthandpressurewavetransmissionspeed.Themethodpresentedinthispapercanbeusedinthedetectionandlocalizationoflittleleakageandslowleakage,iseasytoimplement;moreover,thefalsepositiverateandfalsenegativeratearesignificantlydecrea
7、sed,thepositioningaccuracyisimproved.Thehistoricaldataanalysisverifiestheeffectivenessandfeasibilityoftheproposedmethod.Keywords:oilandgaspipeline;Markovchain;NeymanPearsonanomalydetection;similaritylocation;continuouswaveletlocation气管道故障的有效检测和及时处理已迫在眉睫
此文档下载收益归作者所有