An improved Maxwell-slip model with adaptive control is proposed to reduce the hysteresis nonlinearity of the piezoelectric ceramic actuator, so that it has good hysteresis compensation in a wide frequency band. In classic Maxwell-slip model, the relationship between output force and input displacement will exhibit hysteresis, appearing as a parallelogram, which is close to the hysteresis characteristics of piezoelectric actuators. Since the maximum friction force of each unit slider is proportional to the spring elastic coefficient, if the spring coefficient is a fixed value, then the maximum friction force of each unit is constant in the system real-time control, the output value can be updated by using the adaptive control algorithm, and the input voltage of the piezoelectric actuator can be compensated. To validate the proposed model, a piezoelectric experimental platform was built, and the hysteresis model was used for hysteresis compensation control. The experimental results show that, for the adaptive control of Maxwell-slip model, the root mean square error and mean absolute deviation error in the 0.1~20 Hz wide frequency band are reduced. The displacement error is 0.037 5 μm without control at 0.1 Hz, and the adaptive control method can reduce the positioning error of the piezoelectric micro-positioning platform within 0.012 4 μm. Compared with the classic methods, the proposed adaptive control has the advantages of precision positioning in wide frequency band.