Semi-active fuzzy optimal control of a vehicular multi-dimensional vibration isolator

A multi-dimensional (MD) semi-active vibration isolator for vehicular device is presented in this paper. The isolator is established based on a parallel mechanism with three limbs, each limb includes a sub-system which is composed of a magnetorheological (MR) damper and a linear spring. Then a semi-active fuzzy optimal control method is proposed to reduce the vibration acceleration of the vehicular device. The H∞ controller and the control strategy of MR damper are designed to get the possible output force of the MR damper. As the inverse dynamics of the MR damper is highly nonlinear, the T-S fuzzy model is obtained by genetic algorithm to identify the damper's input current. In addition, a new encoding method is developed when training the fuzzy model to optimize the inputs selection, the rules selection and the antecedentmembership function (MF) parameters simultaneously. Simulation results show that utilizing the presented vibration isolator can significantly reduce the MD vibration response of the vehicular device.