模糊神經(jīng)網(wǎng)絡(luò)在埋地燃氣管道事故后果評價中的應(yīng)用研究

1、Y960609分類號：型122』!密級：——論文編號：Q3Q5Q232貴州大學(xué)2006屆碩士研究生學(xué)位論文模糊神經(jīng)網(wǎng)絡(luò)在埋地燃氣管道事故后果評價中的應(yīng)用研究學(xué)科專業(yè)：材料學(xué)研究方向：管道風(fēng)險評價導(dǎo)師：陶文亮教授研究生：僮祥英中國貴州貴陽2006年5月ABSTRACTBeaucseundergroundgaspipelineanditssubsidiaryfacilitiesarcacomplexsystem，itexitsthedang

2、ersofexplosion、fire、tappingoftoxicity、casualtiesandfacilitiesdestroyedwhenconveyinggas，soitisafrequentphenomenonthatthedisastrousaccidenthappenaccompaniedbytheconveyingoruseofgas。So，itisessentialforeffectivelyevaluatingt

3、heseriousdegreeofpipelinefailuresconsequenceandcombineitwitlltheprobabilityofpipelinefailurestoimproveprecautionarymeasure：toprovidereliablefoundationforscientificdecisionmakingofpipelineriskmanegement：toreasonablelymake

4、useoflimitedresourcesofmanpower、Physicalpower、andfinacialpoweretc：toadoptt11emostadvisablemeasuresforminishingthelosseffectively；toimproveeconomicperformance、protectenviroment，andsoon，andeventuallyhitthetargetofreducingl

5、ossofaccident。，Ihctopicoftheessayis“StudyontheApplicationofFuzzyNeuralNetworkwithCompensationintheEvaluationofSeriousDegreeofUndergroundGasPipelineFailuresConsequence”，itintroduceFuzzyNeuralNetwork，whichisnewbytechnology

6、intotheevaluationofseriousdegreeofundergroundgaspipelinefailuresconsequence，intendedtoestablishanevaluatingsystemwhichbetterreflecttherealperformanceandeffectivelypredicttherisksituationofpipeline，andthenprovidemanagmgfo

7、undationformanagerofpipelineo111esumtotalofstudyintheessay：(1)Throughtheanalyticalstudyofexistingevaluatingtechniqueaboutconsequenceofgaspipelinefailuresbothhereandabroad，analysistheexistedlimitationsofthem，andbasedonthe

8、propegiesoffuzzyneuralnetwork，bringforththemethodofevaluatingtheseriousdegreeofpipelinefailuresconsequencebasedonfuzzyneuralnetwork。(2)Analysisthedynamicalpropertiesofpipelinewhenconveyinggas，bringforwardsimulationofthem

9、ostpossibleaccidentswi血hurtmodel，therebyfigureoutthecasualtiesanddirecteconomiclossesduetoaccident。(3)Collectexistinghistoricaldataaboutfailuresconsequenceofgaspipelinefindoutandclassifyinfluncefactorsoftheseriousdegreeo

10、fpipelinefailuresconsequence，andtransformthemtofuzzylanguagewhichcarlbeusedtotrainingnetworkwithfuzzedtechnique。Makeeveryefforttousetheoriginalandreliabledatatotheutmostextent。(4)Bcauseofthelimitationoforiginaldatarecord

11、，expandtheoriginaldatawithambiguousinformationdiflusedtechnique，iustforobtainmoreusefultainingdttas。Intheprocedure，theinfluenceofinfluncingfactorstoconsequenceevaluationrcstswithfactorweight。BymcansofHierarchialAnalysisc

12、anobtainweiglltofdeterminingaffectingfactors。(5)Trainnetworkandverifyitsfunctionwithpracticalcase，forthepurlzloseofexaminingthereliabilityandaccuracyofnetwork。，(6)Carryoninfluencingexperimentabouttappingoftoxicity。estima