智能交通系統(tǒng)中汽車車型的模糊模式識別.pdf

1、分類號：UDC：5l密級：學號：02074哇東南大掌碩士掌位論文智能交通系統(tǒng)中汽車車型的模糊模式識別研究生姓名：奎塞導(dǎo)師姓名：蔓延皇熬撞申請學位級別蘧論文提交睡贛2塑璺圭基窶學位授予單位盔畝盎掌答辯委員會主席要羹：疊主學科專業(yè)名稱遂纂堂皇撞趔迨論文答辯霞期墊籩蔓i曼豎雷學位授予目明生熙目評閱人&』：盔壘歪2005年1月口箍鑫逢移連德泰壽太學頸：￡學稼論文第～章號富AbatractWimtheEconomicdevelopmentofou

2、rcountrythedemandoftrafficenhancedcontinuouslywhilethetrafficmanagementisrelativelylag，thatmakesthetrafficjamincreasinglyserious，thewaytoimproveitisintroducetheintelligenttransportationsystemInthispaperwedidsomeresearcho

3、fpaRemrecognitionofvehicleautomaticintheintelligenttransportationsystemandpresentedthemethodofestablishrecognitionmodelTherealemanykindsofvehicleinourcountryandtherenodistinetdifferenceamongthem，SOitsverydifficulttogetas

4、atisfiedeffectbyusingtheconventionalrecognitionmethodInthispaperweestablishamethod協(xié)vehiclerecognitionusingthetwostagesandcharacterseparationThecharacterofvehicleisSOmanyandnodistinctdifferenceamongthesestyle，SOinthispape

5、rWeintroducefuzzydegreeforcharacterdistillandusingfuzz)degreeasindexofcharacterforcontributionofclassificationFirststage，wecarryouttheclassificationoftruckandCarbythesupportvectormachinetheorySecondstage，Werespectivelyse

6、tuprecognitionmodelofcarandtruckbyfuzzyneuralnetworks，atthesanletimeweproposedusinggeneticalgorithmoptimizemembershipfunctionandreasoningruleforsolvingadisadvantagethatneuralnetworkseasilygetintolocalminimum，thatmadesyst

7、emshasagoodstabilityandaccuracyInthispaperthecharacterdatawasclassifiedintotraindataandtestdata，thetraindataWasusedtotrainedandmodifiedthemodelparameters，andthetestdatawasusedtotestfiedtheeffectoftherecognitionmodelThere

8、suRofexperimentationtestifythemeanshasvalidityThetheoryofthispaperispropitioustoaccomplishautomaticvehiclerecognitionwhich】ayafoundationforrealizingelectronictollcollectionsystems。Keywords：geneticalgorithm，fuzzyneuralnet