Impedance tracking control of magnetostrictive transducer based on variable stiffness tuning

Abstract

A magnetostrictive transducer (MT) is a highly integrated transducer that enables magneto-mechanical energy conversion when outputting actuation motion or force. To ensure high energy efficiency, an impedance tracking control strategy is proposed based on the variable stiffness tuning mechanism. The transducer's internal impedance is controlled to track the change of the load impedance to obtain the impedance matching for different working conditions, and thus increase the energy efficiency. The developed transducer is driven with magnetostrictive material Terfenol-D. By applying proper magnetic field, the material stiffness is able to be changed and the transducer's internal impedance can be controlled to obtain impedance matching. An impedance network model is built to describe the impedance changing of the transducer. A neural network model is developed to describe the variable stiffness mechanism of Terfenol-D. The impedance tracking is then realized with the feedforward compensation of the inverse neural network model and the inverse impedance network model. The influencing factors of impedance matching have been studied. Simulations and experiments have been conducted to verify the energy efficiency comparisons. Verifications show that the MT efficiency demonstrates big differences with and without the impedance control. The lowest efficiency is 10.8%. With the matching control, the efficiency is 24.7%, and this efficiency is 131% increased.

Keywords

• Magnetostrictive transducer
• Variable stiffness
• Impedance matching
• Neural network model