Modeling and control of magnetorheological fluid dampers using neural networks

Abstract

Due to the inherent nonlinear nature of magnetorheological (MR) fluid dampers, one of thechallenging aspects for utilizing these devices to achieve high system performance is thedevelopment of accurate models and control algorithms that can take advantage of theirunique characteristics. In this paper, the direct identification and inverse dynamic modelingfor MR fluid dampers using feedforward and recurrent neural networks are studied. Thetrained direct identification neural network model can be used to predict the damping forceof the MR fluid damper on line, on the basis of the dynamic responses across the MRfluid damper and the command voltage, and the inverse dynamic neural networkmodel can be used to generate the command voltage according to the desireddamping force through supervised learning. The architectures and the learningmethods of the dynamic neural network models and inverse neural network modelsfor MR fluid dampers are presented, and some simulation results are discussed.Finally, the trained neural network models are applied to predict and controlthe damping force of the MR fluid damper. Moreover, validation methods forthe neural network models developed are proposed and used to evaluate theirperformance. Validation results with different data sets indicate that the proposed directidentification dynamic model using the recurrent neural network can be used topredict the damping force accurately and the inverse identification dynamic modelusing the recurrent neural network can act as a damper controller to generatethe command voltage when the MR fluid damper is used in a semi-active mode.