Title :
MIMO Decoupling Control Based on Support Vector Machines α th-order Inversion
Author :
Fuhua Song ; Ping Li
Author_Institution :
National Key Lab. of Ind. Technol., Zhejiang Univ. Inst. of Ind. Process Control, Hangzhou
Abstract :
Combining inverse system method and support vector machines (SVM), the paper presents an alphath-order inverse system method based on SVM for general nonlinear MIMO discrete systems. The nonlinear offline inverse model of the controlled plant is built by SVM with the radial basis function (RBF) kernel. Cascading SVM alphath-order inversion with the original system can form a pseudo-linear system. So the traditional linear system methods can be used in a nonlinear system. Simulations demonstrate that even without prior information about the plant, the method presented can accurately model the inverse dynamic system of the plant. This method can apply to general nonlinear MIMO discrete systems, and can be easily implemented for engineering
Keywords :
MIMO systems; discrete systems; linear systems; nonlinear control systems; support vector machines; MIMO decoupling control; alphath-order inversion; general nonlinear MIMO discrete systems; inverse dynamic system; inverse system; nonlinear offline inverse model; pseudolinear system; radial basis function kernel; support vector machines; Artificial neural networks; Control systems; Electrical equipment industry; Industrial control; Inverse problems; MIMO; Nonlinear control systems; Nonlinear systems; Process control; Support vector machines; MIMO nonlinear systems; Support vector machines; decoupling; inverse system method;
Conference_Titel :
Intelligent Control and Automation, 2006. WCICA 2006. The Sixth World Congress on
Conference_Location :
Dalian
Print_ISBN :
1-4244-0332-4
DOI :
10.1109/WCICA.2006.1712495