自适应滤波方法研究.pdfVIP

19 7 ( ) V o . 19 No . 7 ( S) 199 8 12 A CT A A ERO N AU T ICA ET A ST R ON A U T I CA SIN ICA Dec. 1998 ( , , 100083) APPROACH TO ADAPTIVE FILTERING ALGORITHM Zhang Chang yun ( D epar tment o f Auto matic Co ntr o , Beij ing U niver sity of A ero nautics and A stro naut ics, Beijing , 100083) Sag e-Hu sa Q R , , R ( Q) ( ) ( ) , Q R Q R , , V 249 - Abstract T his paper pr ov es t hat Sage Husa s adaptive K a man f i tering a go rithm canno t esti- , - mate Q and R simu taneou s y and ana y ses the r eason w hy Sage Husa s a go rithm may ead to f i ter ing div erg ence. A new a g or ithm of adaptive Ka man f i t er is int ro duced w hich can estimat e Q ( or R) accur ate y w hen R ( o r Q) is kno w n . T he most par ticu ar f eat ur e o f this a g or ithm is that o n y matrix mu tip ication and inv ersio n are per for med w hi e no summ atio n o r r eductio n is carr ied out in t he phase o f updating and impro ving the estim atio n of Q( o r R) , hence r esu ting in no f i ter ing div erg ence . Dig it a simu atio n sho w s that it is r emar kab e. Key wordsadaptiv e f i tering , inertia nav ig atio n , integ rat ed navigat ion , [ 1] , Sag e Hu sa , [ 2, 3] , - , , Sag e Hu sa - , Sage Husa 1 - Sage H sa X k + 1 = kX