A directional forgetting algorithm based on the decomposition of the information matrix

被引:57
作者
Cao, LY [1 ]
Schwartz, H [1 ]
机构
[1] Carleton Univ, Dept Syst & Comp Engn, Ottawa, ON K1S 5B6, Canada
关键词
recursive algorithms; least-squares method; parameter estimation; exponential forgetting; directional forgetting;
D O I
10.1016/S0005-1098(00)00093-5
中图分类号
TP [自动化技术、计算机技术];
学科分类号
0812 ;
摘要
A novel algorithm for directional forgetting is proposed based on a matrix decomposition method, which is developed in this paper. This algorithm performs exponential forgetting according to the direction of the data vector, thus preventing the problem known as estimator windup which is a characteristic of the standard exponential forgetting algorithm. (C) 2000 Elsevier Science Ltd. All rights reserved.
引用
收藏
页码:1725 / 1731
页数:7
相关论文
共 12 条
[1]  
Astrom K. J., 1995, ADAPTIVE CONTROL
[2]   CONVERGENCE AND EXPONENTIAL CONVERGENCE OF IDENTIFICATION ALGORITHMS WITH DIRECTIONAL FORGETTING FACTOR [J].
BITTANTI, S ;
BOLZERN, P ;
CAMPI, M .
AUTOMATICA, 1990, 26 (05) :929-932
[3]  
CAO L, 1999, P 7 MED C CONTR AUT
[4]   Rank modifications of semidefinite matrices associated with a secant update formula [J].
Chu, MT ;
Funderlic, RE ;
Golub, GH .
SIAM JOURNAL ON MATRIX ANALYSIS AND APPLICATIONS, 1998, 20 (02) :428-436
[5]  
Hagglund T., 1985, IFAC Proc Vol, V18, P1137
[6]  
Horn R. A., 1986, Matrix analysis
[7]   EXPONENTIAL CONVERGENCE OF RECURSIVE LEAST-SQUARES WITH EXPONENTIAL FORGETTING FACTOR [J].
JOHNSTONE, RM ;
JOHNSON, CR ;
BITMEAD, RR ;
ANDERSON, BDO .
SYSTEMS & CONTROL LETTERS, 1982, 2 (02) :77-82
[8]   STABILIZED LEAST-SQUARES TYPE ADAPTIVE IDENTIFIERS [J].
KREISSELMEIER, G .
IEEE TRANSACTIONS ON AUTOMATIC CONTROL, 1990, 35 (03) :306-310
[9]   RESTRICTED EXPONENTIAL FORGETTING IN REAL-TIME IDENTIFICATION [J].
KULHAVY, R .
AUTOMATICA, 1987, 23 (05) :589-600
[10]  
KULHAVY R, 1984, P 9 IFAC WORLD C BUD, P687