论文部分内容阅读
为了补偿影响压电陶瓷执行器纳米定位系统精度的迟滞非线性,提高系统的控制精度,开展了基于压电陶瓷执行器的迟滞非线性逆模型的研究。兼顾到迟滞的擦除特性和建模的精确度,提出了一种Preisach逆模型分类排序法的神经网络实现方法,用神经网络取代了传统的反查值方法,以避免插值误差。建立三层BP神经网络,运用实测数据进行训练,确定各层权值;然后,结合排序得到的电压和位移极值信息,通过神经网络方法拟合出较精确的输入电压值。运用若干组实验数据检验了此逆模型的有效性,结果表明,该神经网络的实现方法将逆模型的平均误差降低到了1.5V以下,最大误差绝对值降低到了2.7V以下。与反查值方法相比,神经网络实现方法有效提高了压电陶瓷执行器纳米定位系统的迟滞逆模型的精度。
In order to compensate the hysteresis nonlinearity affecting the accuracy of the nano-positioning system of piezoelectric ceramic actuator and improve the control precision of the system, a nonlinear hysteretic nonlinear model based on the piezoelectric ceramic actuator was studied. Taking into account the hysteretic erase characteristics and modeling accuracy, this paper proposes a neural network approach to Preisach inverse model classification and sorting, which replaces the traditional inverse search method with neural networks to avoid interpolation errors. The three-layer BP neural network is established and trained with measured data to determine the weight of each layer. Then, with the help of the voltage and displacement extremum information obtained through the sorting, the accurate input voltage is fitted by neural network. Several sets of experimental data are used to test the validity of this inverse model. The results show that the proposed method reduces the average error of the inverse model to below 1.5V and the maximum absolute error to below 2.7V. Compared with the inverse method, the neural network method can effectively improve the accuracy of the hysteretic inverse model of the PEMA nano-positioning system.