Զѧ
AGTA AUTOMATICA SINICA
1999 25 2 Vol.25 No.2 1999



нģ͵һֽṹʶ1)
ȷ¡Ծ桡
ժҪԾδ֪н(UBB)Իعģͱʶ⣬һµ³ṹѡ񷽷.÷ظ㷨ϢʽֵΪģͶ׼.ͬԡԵļ裬÷ǽ.֤˸÷ǿ.
ؼʡUBBԱʶ³ʶģͶ.
A STRUCTURE IDENTIFICATION METHOD FOR
BOUNDED-ERROR MODELS
SUN Xianfang FAN Yuezu NING Wenru
(Department of Automatic Control, Beijing University of Aeronautics and Astronautics, Beijing100083)
AbstractA new robust structure selection method is proposed to deal with the identification problem of linear regression models with unknown but bounded (UBB) errors. The model-order determination criterion is based on maximizing the determinants' ratio of two ellipsoidal axis information matrixes obtained through repeating a recursive outer-bounding ellipsoid algorithm. Unlike the usual assumptions on noise, namely, independence, constant variance, or martingale difference properties, the assumption in this paper is of asymptotic independence. Strong consistency of the method is proved.
Key wordsUBB error, set membership identification, robust identification, model-order determination.
1
ģͽṹѡϵͳʶһҪɲ.ͳ֪ϵͳģͽṹѡоĽ.ԾUBBϵͳģͽṹѡоԽ.ף1-3ݸ˼ֽṹѡ׼.ף1ǵԱʶòԱаĳһֵʱòǿΪ0Լģ׼.ף2ͨԲвеİԼȷģͽ״ε׼.ף3ԲԱľȷ㷨˶ݻ׼֤¸ѡ񷽷ǿ.ڸ÷Ƚϸӣ³Խϲ4ݣʵñȽ.ף3ݵķΪ˻ڲԱʶظ׵㷨һֽṹѡ׼򣬼Ϣʽֵ׼.㷨һ³ԣ5ݣ򵥣˱㷨Ƕף3㷨ĸĽ.
2ʶ
ģ

(1)
Цǲykǹ۲УkǽͱУekУǷek֪޽Уnǽṹ.k=yk-1yk-nTʱʽ(1)ʾԻع(AR)ģͣʱʽ(1)ʾܿԻع(ARX)ģ.
ϵͳʵֵΪ*ʽ(1)֪κʱkɹ۲ȷһӼ

(2)
ʹæ*Sk.֪йئ*ϢΪ*ʦ0øϢkʱǰĹ۲ݿԵ

(3)
Ѱ򵥣ҾܽذӼk㷨ΪԱʶ㷨.õļԱʶ㷨ظ㷨6ݣ㷨ýһk£

(4)
ʽ(4)ķţף5ݸ.
1.ף6ظ㷨ظ֮һʽ(4).ԦkΪʼظʱѲ(һļ㾫)һؼݻʽ(5)(7)˵һڽ

(5)

(6),(7)
νڽָеϵ

(8)
3ģͽṹѡ׼
ģ͵Ľṹѡ׼֮ǰ¶3ݣ
1.kRnһУһµ

ABȡпѡнBorelƣkǽ.
2.ekĳһʿռ(,F,P)ϵFk=ңei,ikekɵĦҴڳC2C10һУkskʹöκγСh

Уekһβ.
3.kһнУκη׶K=աæ=11++nn,10,,n0det(1,,n)0ʱ

E(kFk-2)C30 a.s.ƣkΪȫλ.
4.kΪһУC4ʹækC4 a.s.,kNƣkΪһн.
壬
1.ʵģͽ״n*nmaxҶÿһѡģͽ״nnmaxУ1)Уekһβģ2)عУkȫλһнģ3)ǽģkʱ 
֤1֮ǰȸһ.
23.ڶ1£nn* a.s.Ϊռnn

(9)
бʾ״Ϊnʱõݻ.
1֤.ʽ(8)2֪nn*ʱ a.s.Ϊռ.

(10)
ʽ(7)֪nn*ʱ

(11)
ʽ(8)1֪

(12)
Knnйص.
ʽ2֪nn*ʱ

(13)
ʽ(11)(13)ɵãkʱмϢʽ׼ǿݵ.
ʽ(13)ڳʼӰ콫ʧϢʽֵ׼Ҳǿݵ.֤.
ԾжṹģͣARXģͣף3ݵķƵĴ.
ڶ1Ե½.
2.ʽ(1)ARģͣekȷֲҶk,0Ϣʽֵ֮ɵģͽ״εǿѡ񣬼ֻҪnmaxn*kʱ
4
ͬף3ݣʽ

УkصľзǶԳܶȵ.βֲܶȺΪ

1ף3׼ͱ׼ѡģͽ״Ϊ123Ƶ(ʵ״Ϊ2)
