基于BP神经网络改进Duhem逆模型的压电陶瓷动态迟滞鲁棒反演滑模控制OA

In order to suppress the effect of dynamic hysteresis in piezoelectric ceramics,a robust inverse sliding mode control strategy based on the feedforward compensation of the BP neural network improved Duhem inverse model recognized by the northern goshawk optimization algorithm is proposed.First,the northern goshawk optimization algorithm is utilized to identify the Duhem model with α,f(·)and g(·).Then,the inverse model of Duhem is derived.Since the nonlinearity of piezoelectric ceramics increases with increas-ing frequency,the output of the Duhem inverse model and the data of actual control voltage and actual displacement are input into the BP neural network for training,so as to obtain a more accurate inverse model.Finally,to compensate the dynamic hysteresis of piezoe-lectric ceramics,a composite control method of robust inverse sliding mode control law combined with inverse model compensation is proposed.The experimental results demonstrate that the accuracy of the proposed method is improved by 3.05％,43.33％,and 29.10％,respectively,compared with a single robust inverse sliding-mode control under tracking signals of different frequencies(10 Hz,40 Hz,and 100 Hz),which verifies the effectiveness of the proposed method.