前馈
控制理论(社会学)
磁滞
振动
非线性系统
蠕动
反馈控制器
反向
振动控制
补偿(心理学)
控制器(灌溉)
计算机科学
材料科学
工程类
控制工程
数学
物理
声学
控制(管理)
几何学
农学
复合材料
精神分析
人工智能
心理学
生物
量子力学
作者
Kam K. Leang,Santosh Devasia
标识
DOI:10.1109/tcst.2007.902956
摘要
In this brief, we study the design of a feedback and feedforward controller to compensate for creep, hysteresis, and vibration effects in an experimental piezoactuator system. First, we linearize the nonlinear dynamics of the piezoactuator by accounting for the hysteresis (as well as creep) using high-gain feedback control. Next, we model the linear vibrational dynamics and then invert the model to find a feedforward input to account vibration - this process is significantly easier than considering the complete nonlinear dynamics (which combines hysteresis and vibration effects). Afterwards, the feedforward input is augmented to the feedback-linearized system to achieve high-precision highspeed positioning. We apply the method to a piezoscanner used in an experimental atomic force microscope to demonstrate the method's effectiveness and we show significant reduction of both the maximum and root-mean-square tracking error. For example, high-gain feedback control compensates for hysteresis and creep effects, and in our case, it reduces the maximum error (compared to the uncompensated case) by over 90%. Then, at relatively high scan rates, the performance of the feedback controlled system can be improved by over 75% (i.e., reduction of maximum error) when the inversion-based feedforward input is integrated with the high-gain feedback controlled system.
科研通智能强力驱动
Strongly Powered by AbleSci AI