Premium
LS‐SVM generalized inverse decoupling control method for multivariable nonlinear discrete systems
Author(s) -
Li Xin,
Li Bo,
Li Ruoqiong
Publication year - 2019
Publication title -
ieej transactions on electrical and electronic engineering
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.254
H-Index - 30
eISSN - 1931-4981
pISSN - 1931-4973
DOI - 10.1002/tee.22900
Subject(s) - multivariable calculus , decoupling (probability) , nonlinear system , control theory (sociology) , inverse system , inverse , mathematics , support vector machine , generalized inverse , computer science , engineering , control engineering , control (management) , artificial intelligence , physics , geometry , quantum mechanics
A new nonlinear decoupling control method without state feedback based on a generalized inverse system of a least squares support vector machine (LS‐SVM) is proposed for a multiple‐input multiple‐output nonlinear discrete system. According to the mathematical model of multivariable nonlinear discrete systems, the reversibility of the LS‐SVM generalized inverse system is analyzed and verified. Using the ability of LS‐SVM to approximate nonlinear functions, the multivariable nonlinear generalized inverse model of stateless feedback can be obtained offline. The generalized inverse system obtained by LS‐SVM identification is connected before the original nonlinear system to decouple the multivariable nonlinear discrete systems into q independent single‐input single‐output pseudolinear subsystems. The theoretical derivation and simulation results show that the presented LS‐SVM generalized inverse decoupling control method can realize the dynamic decoupling of multivariable nonlinear systems. And the pseudolinear composite systems show good dynamic performance. © 2019 Institute of Electrical Engineers of Japan. Published by John Wiley & Sons, Inc.