For the sake of eliminating the hysteresis nonlinearity characteristic and achieving high precision positioning control, a BP neural network with two hidden layers is adopted to establish the hysteresis model, therefore, the relationship between the driving voltage and the output displacement of piezoelectric micro positioning platform can be described and predicted accurately. A feedforward controller based on the inverse hysteresis model is proposed to compensate the hysteresis negative influence and make the positioning system display a excellent linear property. In order to improve the accuracy of the positioning system further, a hybrid control based on the feedforward compensation and expert fuzzy control is designed. The simulation results show that the maximum positioning error is 0.091μm, and the negative effect of hysteresis nonlinearity has been reduced significantly.