压电惯性驱动机构的驱动力和运动参数测试研究

发布时间：2018-02-17 01:24

本文关键词： 压电惯性驱动器驱动力运动参数检测人工神经网络　出处：《浙江师范大学》2015年硕士论文　论文类型：学位论文

【摘要】：压电惯性驱动器是利用压电元件产生的惯性力差作为驱动力的压电驱动装置。压电元件产生的驱动力是机构的动力源,其定量分析对于优化机构的输出性能有着重要作用,驱动装置的运动参数主要包括步长,速度和加速度等,是衡量装置输出特性的重要指标。本文对压电元件产生的驱动力进行了定量分析,设计了驱动力的测试试验和验证试验,将压电元件用作传感元件,设计电路装置测试了驱动力,利用神经网络技术,预测驱动装置的运动参数。研究内容如下：分析压电双晶片元件的输出特性和等效电路模型。推导了正、逆压电效应下压电元件的输出特性,利用Matlab软件仿真分析了压电双晶片元件的尺寸参数对输出位移和端部作用力影响,推导压电元件输出电荷与端部位移关系；分析了压电双晶片元件的阻抗特性,采用阻抗分析仪测试了对压电双晶片元件的等效模型参数。推导压电惯性驱动器的驱动力表达式。分析在正弦信号、锯齿波信号和方波信号激励下压电双晶片的端部位移,速度和加速度响应,分析驱动力的时域和频域响应特性曲线,采用Simulink软件对压电振子动态特性进行了仿真分析。进行压电双晶片振子的驱动力测试试验和验证试验。采用压电式加速度传感器测试了在基频下压电双晶片振子的加速度,结合压电双晶片振子的惯性质量计算得到驱动力的数值,利用快速傅立叶算法分析了驱动力的频域特性；搭建了驱动力验证试验平台,采用摩擦学方法验证驱动力的测试结果。提出基于压电自感知方法的驱动力测试的方法。将压电双晶片的一片压电晶片用作传感元件,分析了压电双晶片的加速度参数与传感元件产生的感应电荷关系,设计了驱动力的测试电路装置。提出利用神经网络技术预测压电惯性驱动器运动参数的预测方法。分析传感元件的感应信号,驱动力和机构运动参数关系,将驱动力测试电路装置的输出信号作为神经网络输入量,构造了用于直线型压电驱动器运动参数预测的BP神经网络模型,实现了运动参数的预测。
[Abstract]:Piezoelectric inertial actuator is a piezoelectric drive device which uses the difference of inertia force produced by piezoelectric element as the driving force. The driving force generated by piezoelectric element is the power source of the mechanism, and its quantitative analysis plays an important role in optimizing the output performance of the mechanism. The motion parameters of the driving device mainly include step size, velocity and acceleration, which are important indexes to measure the output characteristics of the device. In this paper, the driving force produced by the piezoelectric element is quantitatively analyzed, and the driving force test and verification test are designed. The piezoelectric element is used as the sensing element, the circuit device is designed to test the driving force, and the neural network technology is used. The main contents of the study are as follows: the output characteristics and equivalent circuit models of piezoelectric bimorph elements are analyzed. The output characteristics of piezoelectric elements under positive and inverse piezoelectric effects are derived. The effect of dimension parameters of piezoelectric bimorph element on output displacement and end force is simulated by Matlab software, and the relationship between output charge and end position shift is deduced, and the impedance characteristic of piezoelectric bimorph element is analyzed. The equivalent model parameters of piezoelectric bimorph elements were measured by impedance analyzer. The driving force expression of piezoelectric inertial actuator was deduced. The end displacement of piezoelectric bimorph was analyzed under the excitation of sinusoidal signal, sawtooth wave signal and square wave signal. Velocity and acceleration response, time domain and frequency domain response characteristic curves of driving force are analyzed. The dynamic characteristics of piezoelectric oscillator are simulated by Simulink software. The driving force test and verification test of piezoelectric bimorph oscillator are carried out. The piezoelectric accelerometer is used to measure the acceleration of piezoelectric bimorph vibrator at fundamental frequency. Combined with the inertial mass calculation of the piezoelectric bimorph oscillator, the driving force is calculated, and the frequency domain characteristic of the driving force is analyzed by using the fast Fourier algorithm, and the driving force verification test platform is built. The test results of driving force are verified by tribological method. A driving force test method based on piezoelectric self-sensing method is proposed. A piezoelectric wafer of piezoelectric bimorph is used as a sensing element. The relationship between the acceleration parameters of the piezoelectric bimorph and the inductive charge produced by the sensor element is analyzed. The driving force testing circuit device is designed, and the prediction method of piezoelectric inertial actuator motion parameters using neural network technology is proposed. The relationship among sensing signal, driving force and mechanism motion parameter is analyzed. Using the output signal of driving force test circuit as the input of neural network, a BP neural network model is constructed to predict the motion parameters of linear piezoelectric actuator, and the prediction of motion parameters is realized.
【学位授予单位】：浙江师范大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：TN384

【相似文献】