一种基于BP神经网络的压电陶瓷驱动器控制方法研究

Research on a control method of piezoelectric ceramic actuator based on BP neural network

压电陶瓷驱动器在微定位技术中得到了广泛应用,但压电陶瓷的迟滞非线性严重影响了微定位平台的定位精度。因此,为了提高微定位平台的定位精度和响应速度,提出了一种基于反向传播(back propagation,BP)神经网络的压电陶瓷驱动器迟滞非线性补偿方法,以用于进一步对微定位平台控制系统进行研究。首先,通过Prandtl-Ishlinskii(PI)模型确定了神经网络的输入和输出;其次,采集了用于训练神经网络的数据集并对神经网络进行了训练,同时得出了最佳的隐藏层神经元个数;最后,使用训练好的神经网络对压电陶瓷驱动器进行了控制效果测试。实验结果表明,使用该方法来控制压电陶瓷驱动器,可以使其输出位移与理论位移基本相同,验证了神经网络补偿压电陶瓷迟滞非线性的有效性。

Piezoelectric ceramic actuators are widely used in micro-positioning technology,but the hysteresis nonlinearity of piezoelectric ceramics seriously affects the positioning accuracy of micro-positioning platforms.Therefore,in order to improve the positioning accuracy and response speed of the micro-positioning platform,a back propagation(BP) neural network-based hysteresis nonlinearity compensation method for piezoelectric ceramic actuators is proposed for further study of the micro-positioning platform control system.First,the input and output of the neural network are determined by the Prandtl-Ishlinskii(PI) model.Second,the data set for training the neural network is collected and the nural network is trained,and the optimal number of hidden layer neurons is derived.Finally,the control effect of the piezoelectric ceramic actuator is tested using the trained neural network.The experimental results show that using this method to control the piezoelectric ceramic actuator can make its output displacement almost the same as the theoretical displacement,which verifies the effectiveness of the neural network to compensate the piezoelectric ceramic hysteresis nonlinearity.