Identification of MR Damper Based on Normalized Bouc-Wen Model Using Neural Network
The Magneto-rheological (MR) dampers are favorite mechanical system in dynamic structures. This paper presents an application of Wiener-type nonlinear models for describing the hysteresis behaviors of MR dampers at different operating conditions. In this structure, a linear part consisting discrete-time Kautz filters is cascading by a nonlinear mapping function (feedforward neural network (FFNN)). The pole parameters of Kautz filters were chosen with respect to the poles of best fitted linear model on real system. By defining the parameters of Kautz filter, the nonlinear behaviors of system were identified using neural network model, as the output of filters were considered as the output on NN. In order to assess the performances of the developed models a comparison between the responses of the models and another recent modeling approach was preformed.
A. Bahar et al., "Identification of MR Damper Based on Normalized Bouc-Wen Model Using Neural Network", Applied Mechanics and Materials, Vols. 229-231, pp. 2140-2144, 2012