欢迎来到天天文库
浏览记录
ID:36468941
大小:2.10 MB
页数:68页
时间:2019-05-11
《GPS_DR组合定位系统数据融合算法的研究》由会员上传分享,免费在线阅读,更多相关内容在学术论文-天天文库。
1、兰州理工大学硕士学位论文GPS/DR组合定位系统数据融合算法的研究姓名:杨荣荣申请学位级别:硕士专业:控制理论与控制工程指导教师:曹洁20070418GPS/DR组合定位系统数据融合算法的研究AbstractAsallimportantaspectofIntelligentTransportationSystem,vehiclepositioningsystemhasgreatsignificanceinmanyareassuchItSalleviatingn:蛐fk,convenientfordriving,transportat
2、ionmanagement,guardagainsttheRandgivingalarm,urgencyaskingforhelp.AkeypointofVehiclePositioningSystemjstochoosepositioningwaywhichaimstogainaccurateandreliableinformationofvehiclelocation.GPS/DRintegratedpositioningwaynotonlycansolvetheproblemthatsolitaryGPSisunabletOp
3、ositionduetosingalbeingshielded,butcanrestrainthecumulativeerrorofDReffectively.Accuracyandreliabifityofpositioningsystemareimprovedgreatly,SOthiswayhasbeenadoptedwidely.However,forthesakeofcost,cheaperDRsensorsarcadoptedusuallyinGPS/DRintegratedpositioningsystem.Sofus
4、ionalgorithmisindispensabletoimprovetheperformanceofwholesystem.ThatishowtofuselocationinformationofGPSandDReffectively.Therefore,thekeypointofachievingGPS/DRintegratedpositioningisdatafusionscheme,galmanfilteringisabetterway.DatafusionalgorithmofGPS/DRintegratedpositi
5、oningsystemisstudieddetailedbasedonKalmanfilteringtheory.Firstly,filteringmodelofintegratedpositioningsystemisfoundedbasedonvehiclecurrentstatisticalmodel.AppplicationofExtendedKalmanFiltering(EKF),FederatedKalmanFiltering(FKF)andStrongTrackingF.te】ring(STF)inthesystem
6、alestudiedandanalyzedrespectively.WhereasthebaseofthesealgorithmsisEKF,whichhasmanydrawbacksofslowflittingconvergencerote,poorrobustnesstosystemmodelerrorandnoisestatistics,uneasyapplicationinpractice.Inviewofthis,UnscentedKaimanFilteringasanovelnonlinearfilteringmetho
7、disintroduced.Foritisnotn“积卿tolinearizenonlinearsystem,Jacobimatrixcomputingandlinearizationerrorareavioded,SOitspracticabilityandfilteringestimationaccuracyarebetterthan仃aditionalEKEMeantime,accordingtocharacteristicsofGPS/DRintegratedpositioningsystem,UKFissimplified
8、(SUKF)forincreasedcomputationalefficiency.AndFederatedUnscentedKalmanFiltering(Fur,F)aswell雒StrongTrackingUnscentedKa
此文档下载收益归作者所有