提出一种新的率相关迟滞非线性系统的建模方法,并对其在超磁致伸缩作动器建模中的应用进行了研究.与已有的方法比较,所建的模型结构简单.与实验结果对比,模型可以很好地描述作动器对于复合频率输入信号的迟滞非线性.基于模糊树模型,结合神经网络中的逆向学习和专门化学习,提出了一种直接逆模型控制器设计方法.首先离线辨识对象的逆模型作为初始的控制器,然后与对象串联,用LMS算法在线调节控制器中的线性参数.将该方法应用到超磁致伸缩作动器的跟踪控制中,数值仿真结果表明了方法的正确性和可行性. A new method for modeling the rate-dependent hysteresis nonlinear system is proposed and its application in the modeling of Giant Magnetostrictive Actuator is studied.Compared with the existing methods, the proposed model has a simple structure. Compared with the experimental results, the model can well describe the hysteresis nonlinearity of the actuator to the composite frequency input signal.Based on the fuzzy tree model, combined with the reverse learning and specialized learning in the neural network, a direct inverse model controller Design method: Firstly, the inverse model of the object is identified as the initial controller and then connected in series with the object, and the linear parameters in the controller are adjusted online by LMS algorithm.This method is applied to the tracking control of Giant Magnetostrictive Actuator, Numerical simulation results show that the method is correct and feasible.