摘要
针对超精密微位移系统中压电陶瓷驱动器的迟滞非线性问题,提出了一种基于遗传反向传播(BP)神经网络的压电陶瓷迟滞非线性建模方法。通过电涡流位移传感器获取压电陶瓷驱动器不同电压值下所对应的位移值;利用六次多项式拟合获得迟滞的数学模型,从而建立基于遗传BP神经网络的迟滞模型。实验结果显示:该迟滞模型在神经网络测试下的最大误差为0.082 1μm,平均绝对误差为0.015 8μm。表明,所建的迟滞模型能够较精确地反映出压电陶瓷驱动器的迟滞特性,同时为微位移控制系统设计提供了一定的理论基础。
Aiming at problem of hysteresis nonlinearity of piezoelectric ceramic actuator in ultra precise micro displacement system,a nonlinear hysteresis modeling method based on genetic back propagation( BP) neural network is proposed. Get the piezoelectric actuator corresponding displacement values under different voltage values by current eddy displacement sensor; secondly,mathematical model of hysteresis is obtained using six polynomial fitting,so as to establish hysteresis model based on genetic BP neural network. Experimental results show that the maximum error of the hysteresis model is and the average absolute error is. It shows that the hysteresis model can reflect the hysteresis characteristics of piezoelectric actuator accurately,and provides a theoretical basis for the design of micro displacement control system.
引文
[1]韩同鹏,李国平,沈杰.基于压电陶瓷微位移执行器的精密定位技术研究[J].传感器与微系统,2010,29(2):51-53.
[2]范伟,林瑜阳,李钟慎.压电陶瓷驱动器的迟滞特性[J].光学精密工程,2016(5):1112-1117.
[3]范伟,余晓芬.压电陶瓷驱动器蠕变特性的研究[J].仪器仪表学报,2006(11):1383-1386.
[4]张桂林,张承进,赵学良.压电驱动器记忆特性迟滞非线性建模[J].光学精密工程,2012(5):996-1001.
[5]陈增强,郭纯,赵加祥,等.基于Preisach模型的迟滞系统建模与控制[J].控制工程,2006(4):304-306.
[6]Manuel H,Winfried S,Ulrich N.On Maxwell-stefan diffusion in smoothed particle hydrodynamics[J].Heat and Mass Transfer,2016,103:548-554.
[7]冀坤.基于KP模型的磁控形状记忆合金执行器位移控制方法研究[D].长春:吉林大学,2014.
[8]杨斌堂,赵寅,彭志科,等.基于Prandtl-Ishlinskii模型的超磁致伸缩驱动器实时磁滞补偿控制[J].光学精密工程,2013(1):124-130.
[9]于亚婷,杜平安,廖雅琴,等.基于BP网络的电涡流传感器非线性补偿[J].传感器与微系统,2007,26(10):54-56.
[10]江东,单薏,刘绪坤,等.函数拟合法力数字传感器的非线性和温度补偿[J].传感器与微系统,2016,35(2):16-18.
[11]Alvise S,Marco V.Polynomial fitting and interpolation on circular sections[J].Applied Mathematics and Computation,2015,258:410-424.
[12]钱飞,许素安,刘亚睿,等.基于多项式拟合的压电陶瓷迟滞神经网络建模[J].计算机仿真,2015(1):361-366.
[13]Elsa P,Domin P,Guil M.Dynamic hysteresis modelling of entangled cross-linked fibres in shear[J].Sound and Vibration,2016,383(24):248-264.
[14]Zhou Y C,Huang F Z,Cheng Y B.Numerical analysis of a hysteresis model in perovskite solar cells[J].Computational Materials Science,2017,126:22-28.
[15]杨磊,韩邦成,孙津济.基于ANSYS有限元法的电涡流位移传感器分析[J].传感器与微系统,2007,26(10):15-17.