Decomposition strategy-based hierarchical least mean square algorithm for control systems from the impulse responses

In this research, the issue of parameter estimation for control systems is considered to develop a highly efficient estimation approach for the purpose of satisfying the need of industrial process modelling. For dynamical production processes, an error objective function in accordance with the dynam...

Full description

Saved in:
Bibliographic Details
Published in:International journal of systems science Vol. 52; no. 9; pp. 1806 - 1821
Main Authors: Xu, Ling, Ding, Feng, Zhu, Quanmin
Format: Journal Article
Language:English
Published: London Taylor & Francis 04.07.2021
Taylor & Francis Ltd
Subjects:
Algorithms
Control system
Control systems
Decomposition
hierarchical identification
least mean square
Mean square values
Parameter estimation
recursive algorithm
ISSN:0020-7721, 1464-5319
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:In this research, the issue of parameter estimation for control systems is considered to develop a highly efficient estimation approach for the purpose of satisfying the need of industrial process modelling. For dynamical production processes, an error objective function in accordance with the dynamically sampled data is constructed for on-line identification. In order to simulate the instantaneous response of dynamical processes, the experimental scheme of impulse responses is adopted, and the observational data of impulse responses are used as the identification experimental data. In order to acquire high accuracy and stable performance, a hierarchical least mean square method is designed by means of the decomposition technique and the hierarchical principle. Finally, the superiority of the hierarchical least mean square approach is verified by the comparison simulation experiment and the effectiveness of the hierarchical least mean square method is proved by the detailed numerical examples.
Bibliography:ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 14
ISSN:0020-7721
1464-5319
DOI:10.1080/00207721.2020.1871107