In order to realize high precision tracking control of the Giant Magnetostrictive Administer(GMA)

a numerical compensation approach was presented based on inverse Preisach model and a series of optimized experimental methods were discussed. The numerical expressions of the classical Preisach model were presented in detail for different input variations and the inverse classical Preisach model was established. The First Order Descending(FOD) datasets were obtained from the identification experiments and smoothed by a FFT filter. A good agreement between measured hysteresis loops and predicted curves shows that the FFT filter is an effective tool to cancel the errors and noises of the FOD datasets. In the tracking experiments

the Preisach inverse model was used to compensate the hysteresis of GMA to obtain a good tracking performance. On a moving range of 0~34 μm

the tracking error of the GMA with compensation is less than -2.9%~+2.7%

compared with the error of -14.7%~+11.2% without compensation. The experimental results indicate that the compensation approach based on numerical inverse Preisach model can greatly improve tracking performance of GMA

and the optimized experimental methods are effective.