自适应卡尔曼滤波算法
AnoptimaladaptiveKalman lter179
x¯kshouldbechangedinto(1/αk) x¯k,i.e.,thecovariance
¯kisenlarged.Frommatrixofthepredictedstatevectorx
thediscussionabove,anoptimaladaptivefactorisobtainedbasedonthefollowingtheorem.
Theorem1Assumethatanestimateofthecovariancema-¯kisobtainedandde-trixofthepredictedresidualvectorv
vnotedas ¯k,andatheoreticaloneoutputfromtheadaptive v lteringis ¯k,thenanoptimaladaptivefactorshouldsatisfy
thecondition v v(13) ¯k= ¯k.TheoptimaladaptivefactorbasedonEq.(13)is
tr( v¯k k)αk=.
vtr( )¯kk
Theremainingproblemistoevaluatetheestimatedcovari- vancematrix ¯k.Assumethatthepredictedresidualvector
¯k)=0;thusthecovariancematrixofv¯kcan¯ksatis esE(vv
beexpressedas¯kv¯T v¯k=E(vk).
(23)
Therearetwostrategiestoevaluate v¯k.
Strategy1:EstimatorbasedonSage’swindowing.
¯kisanaverageofTheestimatedcovariancematrixofv
T¯k iwithina xedwindowN,¯k ivv v ¯k
1
¯k iv¯T=vk i.N 1
i=1N
(24)
(14)
ProofInadaptive ltering,thecovariancematrix x¯kis
changedto(1/αk) x¯k,thentheestimatedcovariancematrix v ¯kofthepredictedresidualvector,similartoEq.(12),is1T v Ak x¯k=¯kAk+ k.
kConsideringEq.(13),wehave v ¯k
1T
=Ak x¯kAk+ k.k
(16)(15)
Strategy2:Estimatorbasedonthepredictedresidualvectoratthepresentepoch.
¯k iInordertoavoidthesmoothnessoftheaverageofv
T¯k iwithinNepochs,whichdoesnotaccuratelyre ectthev
uncertaintyofthemodelerrorsattk,theestimatedcovariance
¯kcanbeestimatedbyv¯kv¯Tmatrixofvkatthepresentepoch,
whichis v¯kv¯T ¯k=vk.
TakingthetraceofbothsidesofEq.(25),wehave v¯kv¯T¯T¯k=tr( ¯k)=tr(vk)=vkv
nk
i=1
2v¯k,i
(25)
MultiplyingαkinbothsidesofEq.(16),wearriveat
T vαk ¯k=Ak x¯kAk+αk k
(26)
(17)(18)
or
T vαk( ¯k k)=Ak x¯kAk.
wherenkmeasurementsarecollectedateachepochtk.From
Eqs.(26)and(14),wearriveatacorollary.
Corollary1Inpracticalapplication,theoptimaladaptivefactorbasedontheestimatedcovariancematrixofthepre-dictedresidualvectoris
nk
22¯ki σi)i=1(σv
αk=k(27) σ)(v¯ii=1kiandtheapproximatelyoptimaladaptivefactoris
v tr( 1¯k)<tr( v¯k) nk2
αk=¯ki=1σv
i otherwise, v¯i=1ki
Adding ktobothsidesofEq.(18)andconsideringEq.(12),
wehave
vαk( (19)¯k k)+ k= v¯kor
vαk( ¯k k)= v¯k k.
(20)
Takingthetraceoftheaboveexpression,thenweobtaintheoptimaladaptivefactorinEq.(14),whichcompletestheproof. Inpracticalapplications,weusuallyhopethattheadap-tivefactorαk≤1,thentheoptimaladaptivefactorcanbeexpressedas
v 1tr( ¯k)<tr( v¯k)
)αk=tr( v(21)¯k
otherwise. ( v¯k k)Furthermore,thenumeratoranddenominatorofEqs.(14)or
(21)containthesamecovariancematrix k.Omittingit,wegetanapproximatelyoptimaladaptivefactoras
v 1tr( ¯k)<tr( v¯k))αk≈tr( v(22)¯k
otherwise. ( v¯k)
(28)
2inwhichσv¯kcanbeexpressedas
i
2T2
σv¯kaki+σi¯k=aki x
i
(29)
2whereakiistheithrowofAk.Itissimpletocalculateσv¯kibyusingEq.(29).
Analyzingthederivationabove,we ndthat:
v1.If ¯k= v¯k,thenαk>1,andtheadaptive lterdegen-eratesintothestandardKalman lter.2.If
vtr( ¯k k)>tr( v¯k k)
(30)
The purpose of an adaptive filter is to reduce or bound the gaps by ...Another approach is to estimate the optimal Kalman gain directly without ...
An active filter based o... 4页 1财富值 An Adaptive Color-Based ... 22页 免费 An optimal adaptive Kalm... 7页 5财富值 Adaptive Fading Kalman ...
An optimal adaptive Kalm... 7页 5财富值 Adaptive Fading Kalman F... ... [95.06]Steering an adaptive antenna array by the simplified Kalman filter...
Adaptive Kalman Filter o... 暂无评价 5页 2.00元 Blind Channel ...In addition, particle filtering as an optimal nonlinear/non-Gaussian filtering...
(or in probability)an optimal estimator for H{z.in an oKFs.Then,these ...(optimized multiple model adaptive Kalman filter, oMM—AKF);由于oMM—AKF算法...
文章编号:1001‐2400(2011)06‐0119‐07 An Algorithm of Adaptive Kalman Filter in AHRS TIAN Yi,SUN Jinhai,LI Jinhai,YAN Yuepeng (Institute of ...
An Adaptive Color-Based Particle Filter_专业资料...mean shift or the Kalman/mean shift tracker the...Ramesh, Mean Shift and Optimal Prediction for E?...
The federated filter yields estimates that are globally optimal or ...Neural network-aided adaptive unscented Kalman filter for nonlinear state ...
an adaptive Kalman filter. The simulation demonstrated that both methods could improve the performance of the filter. However,considering of practical limits ...
Kalman滤波和MATLAB实现_信息与通信_工程科技_专业资料。卡尔曼滤波器(Kalman Filter)是一个最优化自回归数据处理算法(optimal recursive data processing algorithm)...
热门文档