Many asymmetric hysteresis models based on traditional BoucWen for the piezoelectric actuator have some redundant parameters, which reduce the accuracy of parameter identification. The most frequently used particle swarm algorithm converges slowly and is easy to fall into local optimum in terms of parameter identification of the piezoelectric actuator. Thus, a normalized asymmetric hysteresis model is proposed by introducing two polynomials to describe the asymmetric hysteresis behavior and using the normalized BoucWen model to eliminate the redundancy of the parameters. Tthe selfadaptive differential evolution algorithm is developed for parameter identification, in which both associated control parameters and trial vector generation strategies can be selfadapted with the increase of generations. An experimental system about the piezoelectric actuator is set up. The results show that the proposed model is better to represent the actual characteristic of the piezoelectric actuator and successfully eliminate the redundancy of the parameters, which decreases the difficulty of parameter identification. Compared with traditional differential evolution algorithm and particle swarm algorithm, the selfadaptive differential evolution algorithm can find the optimal solution more quickly and more accurately.