A coupled recursive least squares algorithm for multivariable systems and its computational amount analysis by using the coupling identification concept

Summary In order to solve the problem of the parameter identification for large‐scale multivariable systems, which leads to a large amount of computation for identification algorithms, two recursive least squares algorithms are derived according to the characteristics of the multivariable systems. T...

Celý popis

Uloženo v:
Podrobná bibliografie
Vydáno v:International journal of adaptive control and signal processing Ročník 38; číslo 2; s. 513 - 533
Hlavní autoři: Jin, Yu, Ding, Feng
Médium: Journal Article
Jazyk:angličtina
Vydáno: Bognor Regis Wiley Subscription Services, Inc 01.02.2024
Témata:
Algorithms
Coupling
coupling identification concept
Least squares
multivariate system
Parameter estimation
Parameter identification
recursive identification
Subsystems
ISSN:0890-6327, 1099-1115
On-line přístup:Získat plný text
Tagy: Přidat tag
Žádné tagy, Buďte první, kdo vytvoří štítek k tomuto záznamu!
Popis
Shrnutí:Summary In order to solve the problem of the parameter identification for large‐scale multivariable systems, which leads to a large amount of computation for identification algorithms, two recursive least squares algorithms are derived according to the characteristics of the multivariable systems. To further reduce the amount of computation and cut down the redundant estimation, we propose a coupled recursive least squares algorithm based on the coupling identification concept. By coupling the same parameter estimates between sub‐identification algorithms, the redundant estimation of the subsystem parameter vectors are avoided. Compared with the recursive least squares algorithms, the proposed algorithm in this article have higher computational efficiency and smaller estimation errors. Finally, the simulation example tests the effectiveness of the algorithm.
Bibliografie:ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 14
ISSN:0890-6327
1099-1115
DOI:10.1002/acs.3712