不确定观测系统的分布式状态估计

《不确定观测系统的分布式状态估计》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、AbstractInthepractice，thereexistmanystochasticsystemswithuncertainobservations．Forexample，duetotheeffectsofenvironmentandunreliablecommunicationnetworks，datalossmayoccureinthenetworkedcontrolsystemsandsensornetwork．Thustheuncertaintyappearsinmeasurementequation．Inthispaper，westudythesyst

2、emswithdatalosswhichcanbedescribedbyaBernoullidistributedrandomvariable．Formulti-sensordiscrete-timestochasticlinearsystems、Ⅳitlluncertainobservations，westudythedistributedinformationfusionestimationalgorithms．硼1emaincontentsincludethedistributedinformationfusionwhitenoiseestimators，stat

3、eestimatorsandstateestimatorswithunknowildisturbancesinsensorobservations．Basedontheinnovationanalysismethod，thelinearminimumvarianceoptimalwhitenoiseestimators，includinginputwhitenoiseestimatorsandobservationwhitenoiseestimators，arepresentedforsingle—sensordiscrete—timestochasticsystems

4、withuncertainobservationsandcorrelatednoises．Andthesteady-statewhitenoiseestimatorsaregivenforthestablesystems．Furthermore，theunifiedasymptoticallystableWienerwhitenoiseestimatorsaleobtainedbasedonthesteady-stateKalmanestimators．Finally，basedonthedistributedoptimalweightedfusionestimatio

5、nalgorithmsinthelinearr蛐unlvariancesense，thedistributedweightedfusioninputwhitenoiseestimatorsareproposedformulti—sensorsystemswithuncertainobservations．Theestimarionerrorcross—covariancematrixbetweenanytwosensorsubsystemsisderived．Tocompare、^，itllthecentralizedfusionestimatorsonaccuracy

6、andcomputationalcost，thecentralizedfusionestimatorsarealsogiven．Basedontheinnovationanalysismethod，thelinearmini／nunlvarianceoptimalandsteady-statestateestimatorsarepresentedforsingle··sensordiscrete-timestochasticlinearsystemswithuncertainobservationsandcorrelatednoises．Furthermore，theu

7、nifiedasymptoticallystableWienerstateestimatorsareobtainedbasedonthesteady-stateKalmanfilter．Inaddition，thestateestimatorsaredevelopedforasingle—sensorstochasticsystem、^，iⅡluncertainobservationsandcorrelatednoisesbothatthesametimeandtheneighborhoodtime．Furthermore，thecros