資源描述:
《基于模糊時間序列的傳感器網(wǎng)絡(luò)感知數(shù)據(jù)預(yù)測模型》由會員上傳分享,免費在線閱讀,更多相關(guān)內(nèi)容在學(xué)術(shù)論文-天天文庫。
1、第22卷第2期運籌與管理Vol.22,No.22013年4月OPERATIONSRESEARCHANDMANAGEMENTSCIENCEApr.2013基于模糊時間序列的傳感器網(wǎng)絡(luò)感知數(shù)據(jù)預(yù)測模型南國芳,周帥印,李敏強,寇紀(jì)淞(天津大學(xué)系統(tǒng)工程研究所,天津300072)摘要:傳感器網(wǎng)絡(luò)監(jiān)控系統(tǒng)屬于大型復(fù)雜系統(tǒng),由感知節(jié)點以一定的時間間隔向sink節(jié)點發(fā)送感知數(shù)據(jù),以實現(xiàn)對應(yīng)用環(huán)境的監(jiān)控。由于網(wǎng)絡(luò)本身及應(yīng)用環(huán)境的影響,得到的感知數(shù)據(jù)往往存在不確定性。此外,周期性報告數(shù)據(jù)模式影響到實時監(jiān)控數(shù)據(jù)的精確性。本文應(yīng)用時間序列模型預(yù)測傳感器數(shù)據(jù)以響應(yīng)用戶查詢,可有效降低網(wǎng)絡(luò)通信量
2、。通過對無線傳感器網(wǎng)絡(luò)的數(shù)據(jù)分析,引入多屬性模糊時間序列預(yù)測模型,充分考慮了無線傳感器網(wǎng)絡(luò)時間序列中存在的趨勢因素,并提出了適合于傳感器網(wǎng)絡(luò)的修正預(yù)測模型。實驗結(jié)果表明模糊時間序列模型可有效預(yù)測傳感器網(wǎng)絡(luò)數(shù)據(jù),且能提高預(yù)測精度。關(guān)鍵詞:信息管理與信息系統(tǒng);模糊時間序列預(yù)測模型;傳感器網(wǎng)絡(luò);分布式數(shù)據(jù)庫中圖分類號:TP18文章標(biāo)識碼:A文章編號:1007-3221(2013)02-0143-07ForecastingModelofSensingDataforSensorNetworksBasedonFuzzyTimeSeriesNANGuo-fang,ZHOUShuai
3、-yin,LIMin-qiang,KOUJi-song(InstituteofSystemsEngineering,TianjinUniversity,Tianjin300072,China)Abstract:Themonitoringsystembyusingsensornetworksbelongstolargescaledcomplexnetwork,wheresens-ingdataaredeliveredatafixedtimeintervalinapredefinedwaytothesinknodeforuserqueries,therebytheappl
4、icationenvironmentcanbemonitored.Duetothefactthatthequalityofserviceisaffectedbysensornet-worksandtheapplicationenvironment,thesensordatacollectedisusuallywithuncertainty.Inaddition,themechanismofperiodicallyreportmayalsoleadtoinaccurateinformationforrealtimeapplication.Inthispaper,weap
5、plytimeseriesmodeltoforecastthesensordata,thenresponsetheuserqueries,whichwillreducethecommunicationoverhead.Byanalyzingsensingdataproducedbysensornetworks,amulti-attributefuzzytimese-riesforecastingmodelisalsointroduced,andittakesthetrendfactorexistingintimeseriesintoconsideration.Anim
6、provedmodelthatsuitstheforecastofsensingdataisultimatelyproposed.Simulationresultsshowthattheproposedfuzzytimeseriesforecastingmodelcaneffectivelypredictfuturesensingdataofsensornetworksandim-provethepredictingaccuracy.Keywords:informationmanagementandinformationsystem;fuzzytimeseriesfo
7、recastingmodel;sensornet-works;distributeddatabase0引言面向工業(yè)生產(chǎn)監(jiān)控的移動無線傳感器網(wǎng)絡(luò)監(jiān)控系統(tǒng)屬于大型復(fù)雜系統(tǒng),是由大量的同質(zhì)或異質(zhì)的移動感知節(jié)點以自組網(wǎng)的形式構(gòu)成,其目的是協(xié)同地感知、采集和處理網(wǎng)絡(luò)覆蓋區(qū)域中感知對象的信息,從而實現(xiàn)對復(fù)雜環(huán)境的監(jiān)控。近年來,無線傳感器網(wǎng)絡(luò)受到了研究者越來越多的重視,且被應(yīng)用到許多不同[1~3][4][5]的領(lǐng)域中,例如農(nóng)業(yè),環(huán)境監(jiān)控,工業(yè)監(jiān)控等。目前,無線傳感器網(wǎng)絡(luò)節(jié)點主要被部署在很多敏收稿日期:2011-12-03基金項目:國家自然科學(xué)基金資助項目(7107