项目名称: 压电执行器速率依赖迟滞特性的非线性内模控制研究
项目编号: No.61304134
项目类型: 青年科学基金项目
立项/批准年度: 2014
项目学科: 自动化技术、计算机技术
项目作者: 邓亮
作者单位: 上海电力学院
项目金额: 23万元
中文摘要: 压电执行器由于定位精度高、驱动力大、快速响应等特点,而广泛地应用于超精密定位平台。然而其固有的迟滞特性,给精密定位系统的控制带来困难。本项目以精密定位系统中压电执行器的控制为背景,研究变频输入下,压电执行器速率依赖的迟滞特性的非线性内模控制理论与方法。主要内容包括:建立以迟滞算子为外生输入的非线性滑动平均模型串联自回归滑动平均模型构成的伪Hammerstein模型描述速率依赖的迟滞特性,并研究其参数的辨识方法。针对速率依赖的迟滞难于求逆,提出一种改进的逆间隙算子,并构造以此算子为隐层神经元的改进的Elman神经网络描述速率依赖的逆迟滞特性。基于所辨识的伪Hammerstein模型并以此Elman神经网络为初始内模控制器,研究压电执行器基于输入到状态稳定的非线性内模控制策略。本项目对压电执行器在微纳米定位技术领域的进一步研究和发展具有重要的理论意义和实际应用价值。
中文关键词: 速率依赖的迟滞;伪Hammerstein模型;Elman神经网络;内模控制;压电执行器
英文摘要: For the high positioning accuracy,the large driving forces and the rapid response capability,the piezoelectric actuators are widely applied to ultra-precision positioning stages.However, the inherent hysteresis makes the precision positioning system difficult to control.Based on the background of the control of piezoelectric actuators in precision positioning systems, this project investigates the theory and method of the nonlinear internal model control of rate-dependent hysteresis characteristics in piezoelectric actuators under the excitation of the input with varying frequency.The main contents include: the pseudo-Hammerstein model, comprised of a nonlinear moving average model with a hysteretic operator as the exogenous input in series with an auto-regressive moving average model,is established to describe the rate-dependent hysteresis characteristics and the identification approach of the parameters is studied. Considering the problem that it is difficult to derive the inverse of rate-dependent hysteresis, a kind of modified inverse backlash operator is proposed and is taken as the neuron of the hidden layer of the modified Elman neural network to characterize the rate-dependent inverse hysteresis. Based on the identified pseudo-Hammerstein model and the initial internal model controller constructed by the
英文关键词: rate-dependent hysteresis;pseudo-Hammerstein model;Elman neural network;internal model control;piezoelectric actuator