Modeling and control for nonlinear structural systems via a NN-based approach

In this study, we present a neural network (NN) based approach which combines H(alpha) control performance with Tagagi-Sugeno (T-S) fuzzy control for use in nonlinear structural systems. The NN model is adopted to deal with the modeling errors of nonlinear structural systems under external excitation. Fuzzy-model-based H(alpha) control is designed by means of linear matrix inequality (LMI) methods as derived from the Lyapunov theory. A tuned mass damper is designed on a nonlinear structural system where the first frequency mode is utilized to reduce the state response under external resonant disturbances. Then the feedback gain of the said fuzzy controller needed to stabilize a nonlinear structural system is calculated using the Matlab LMI toolbox, The proposed method is then applied to a nonlinearly tuned mass damper system. The simulation results show that not only is the proposed method able to stabilize a nonlinear structural system, but also has strong robustness in terms of preventing modeling errors and external excitations. (C) 2008 Elsevier Ltd. All rights reserved.