Fuzzy indirect adaptive control using SVM-based multiple models for a class of nonlinear systems

• Published in: Neural computing & applications Vol. 22; no. 3-4; pp. 825 - 833
• Main Author: Zhou, Yonghua
• Format: Journal Article
• Language: English
• Published: London Springer-Verlag 01.03.2013; Springer
• Subjects: Applied sciences; Artificial Intelligence; Computational Biology/Bioinformatics; Computational Science and Engineering; Computer Science; Computer science; control theory; systems; Data Mining and Knowledge Discovery; Exact sciences and technology; Image Processing and Computer Vision; Learning and adaptive systems; Original Article; Probability and Statistics in Computer Science; Indirect adaptive control; Fuzzy control; Adaptive learning; Multiple models; Support vector machine; Nonlinear system; Neural computation; Uncertainty; Identification; Neural network; Controller; Fuzzy set; Non linear system; Adaptive control; Modeling; State variable; Learning; Kernel function; Behavior; Numerical stability; Adaptation
• ISSN: 0941-0643; 1433-3058
• DOI: 10.1007/s00521-012-1313-7

Adaptive control with multiple models can further improve the adaptation ability of controllers for the plant with wide-range uncertain parameters. Fuzzy modeling and control are introduced into the multiple-model adaptive control in this paper, which facilitates the intelligent behavior of a plant facing with uncertainty. Within the combination of fuzzy sets of state variables, the corresponding combined kernel functions of support vector machine are utilized to describe the unknown nonlinear dynamics. The coefficients of kernel functions are learned online through adaptive laws. The multiple identification models and indirect adaptive controllers are assigned to the plant through fuzzy inferences. The stability of adaptive law corresponding to the fuzzy identification model and the synthetic control input through fuzzy fusion has been proved for the proposed fuzzy multiple-model adaptive control (FMMAC). The simulation results demonstrate that the proposed FMMAC can achieve favorable control performance for a class of nonlinear systems.