DocumentCode :
1887993
Title :
Nonlinear Control System Intelligent Identification Using Optimized Support Vector Machines
Author :
Zhu, Jia-yuan ; Zhou, Hong ; Huang, Xian-cong ; Li, Mao-hui
Author_Institution :
Dept. of Land-based Early-Warning Surveillance, Air Force Radar Acad., Wuhan, China
fYear :
2010
fDate :
25-26 Dec. 2010
Firstpage :
1
Lastpage :
3
Abstract :
Nonlinear control system identification is studied using neoteric optimized Least Squares Support Vector Machines (LS-SVM) in this paper. Firstly, a multi-layer adaptive optimizing parameters algorithm is developed for improving learning and generalization ability of least squares support vector machines. According to different learning problems, the optimization approach can obtain appropriate LS-SVM parameters adaptively. Then, a nonlinear control system is identified by improved LS-SVM. The results show that the optimization approach can acquire best-optimized parameters for LS-SVM, and optimized LS-SVM can provide excellent control system identification precision and excellent convergence. And also, the multi-layer adaptive optimizing parameters algorithm may be appropriately extended to other types of support vector machines.
Keywords :
identification; least squares approximations; nonlinear control systems; support vector machines; control system identification precision; multilayer adaptive optimizing parameters; neoteric optimized least squares support vector machines; nonlinear control system identification; nonlinear control system intelligent identification; optimization approach; optimized support vector machines; Artificial intelligence; Estimation; Kernel; Nonlinear control systems; Optimization; Support vector machines;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Information Engineering and Computer Science (ICIECS), 2010 2nd International Conference on
Conference_Location :
Wuhan
ISSN :
2156-7379
Print_ISBN :
978-1-4244-7939-9
Electronic_ISBN :
2156-7379
Type :
conf
DOI :
10.1109/ICIECS.2010.5677784
Filename :
5677784
Link To Document :
بازگشت