A moving-bank based intelligent control

A moving-bank based intelligent control method was introduced to deal with the complex nonlinear system. Firstly, the predictive model of the nonlinear system was modeled online based on the least squares support vector machine (LS-SVM), and the linearized predictive model was acquired. Secondly, the multiple fixed model bank was dynamically updated based on the linearized predictive model. Finally, a moving-bank was constructed by choosing the models which near the operating region from the fixed model bank, according to a choosing mechanism; and controllers were reconfigured accordingly based on the moving-bank. The simulation results demonstrate the effectiveness of this method.

[1]  Johan A. K. Suykens,et al.  Least Squares Support Vector Machine Classifiers , 1999, Neural Processing Letters.

[2]  Bokyung Jung,et al.  Multiple Model Adaptive Control Scheme for Nonlinear Reconfigurable System , 2006, 2006 SICE-ICASE International Joint Conference.

[3]  P. S. Maybeck,et al.  Flexible spacestructure control via moving-bank multiple model algorithms , 1994 .

[4]  J. Suykens.  Nonlinear modelling and support vector machines , 2001, IMTC 2001. Proceedings of the 18th IEEE Instrumentation and Measurement Technology Conference. Rediscovering Measurement in the Age of Informatics (Cat. No.01CH 37188).

[5]  G. Ducard,et al.  A reconfigurable flight control system based on the EMMAE method , 2006, 2006 American Control Conference.

[6]  Peter S. Maybeck,et al.  Space structure control using moving bank multiple model adaptive estimation , 1988, Proceedings of the 27th IEEE Conference on Decision and Control.