Aiming at the problem that the traditional Bouc-Wen model cannot accurately characterize the intrinsic hysteresis asymmetry of piezoelectric actuators thus caused the limited control precision, in this paper, a hysteresis modeling method based on the combination of traditional Bouc-Wen model and Hammerstein model is proposed. The control strategy based on the sliding mode control using Kalman state estimation is used to improve the control accuracy. First, the parameters of the hysteresis model combined with the traditional Bouc-Wen model and the Hammerstein structure are identified by using particle swarm optimization (PSO). Then, the Kalman observer is built based on the state space matrix of the second-order system. According to the state prediction of Kakman observer, the feedback compensation is made by using sliding mode law and the feedforward compensation is made by establishing the inverse model, forming a feedforwar-feedback composite compensation. The numerical simulation shows that the nonlinear model can well describe and predict the dynamic output of piezoelectric ceramic actuator in the range of 0-100 V peak value and 0-100 Hz excitation frequency. The displacement of the actuator is between 0 and 6 μm, the average hysteresis error of the traditional Bouc-Wen model is 0.654 95 μm and 0.186 39 μm, respectively, at 50 Hz, based on the traditional Bouc-Wen model and Hammerstein structure, the room mean square (RMS) error of feedforward proportion integration differentiation (PID) compensation is 0.088 5 μm, and the RMS error of feedforward sliding mode composite control compensation is 0.047μm, which is only 0.78% of the maximum output displacement, the maximum tracking error is only 0.153 μm, and the precision is improved by 78%, indicating that the proposed hysteresis model based on traditional Bouc-Wen model and Hammerstein structure and its compensation control algorithm are helpful to realize high-speed and wide-frequency ultra-precision positioning control of piezoelectric ceramic actuator.