Recursive information forgetting with augmented UD identification
作者:
STEVES. NIU,
D.GRANT FISHER,
期刊:
International Journal of Control
(Taylor Available online 1996)
卷期:
Volume 63,
issue 3
页码: 623-637
ISSN:0020-7179
年代: 1996
DOI:10.1080/00207179608921861
出版商: Taylor & Francis Group
数据来源: Taylor
摘要:
Fundamental analysis of information forgetting schemes in recursive identification shows that they can be classified into two basic approaches: relative and absolute forgetting. The augmented UD identification (AUDI) algorithm developed by the authors reveals that all the information pertinent to identification and information forgetting is contained in a single matrix called the augmented information matrix (AIM). The tracking ability of the recursive algorithms is determined by the relative weight that is put on to the new information contained in the regressor, compared with the information already contained in the AIM. This relative weight can be easily controlled by the simple diagonal loss function matrix produced by the AUDI method. This simplifies the interpretation of both new and existing information forgetting approaches and also provides helpful guidelines for design.
点击下载:
PDF (573KB)
返 回