摘要
由于迭代学习前馈反馈控制方法面对具有重复运动性质系统时,既能加快收敛速度又能降低收敛误差,本研究将其引入到微纳操纵成像系统来提高扫描器的跟踪精度。首先,提出了开环比例微分(Proportional differential,PD)型迭代学习结合闭环反馈PD型学习律,并给出了学习律收敛条件,同时为了降低系统运行时间,提高学习效率,将传统的固定学习增益变为指数变增益。其次搭建了基于微纳操纵成像系统的迭代学习控制器,并进行了仿真分析。结果表明,相较于开环迭代学习控制、闭环迭代学习控制,迭代学习前馈反馈控制最大收敛误差最低,且鲁棒性强,算法易于实现,能有效地满足扫描时轨迹跟踪的精度要求。
Since the feedforward and feedback control method foriterative learningcan accelerate convergence speed and reduce convergence error for systems that have repetitive motion properties, itis introduced intothe micro-nano manipulation imaging system to improve the tracking accuracy of a scanner. Firstly, an open-loop PD type iterative learning law is combined with a closed-loop feedback PD type learning law, and the convergence condition of the learning law is given. At the same time, in order to reduce the operating time of themicro-nano manipulation imaging system and improve its learning efficiency, the traditional fixed learning gain is changed into the exponential variable gain. Secondly, an iterative learning controller based on the micro-nano manipulation imaging system is built and simulated. Thesimulation results show that, compared with the open-loop iterative learning control and the closed-loop iterative learning control, the maximum convergence error is low and the robustness is strong.The method is easy to implement and effectively meets the requirements for trajectory tracking accuracy during scanning.
作者
吴文鹏
王一帆
胡贞
WU Wenpeng;WANG Yifan;HU Zhen(School of Electronic Information Engineering,Changchun University of Science and Technology,Changchun 130022,China)
出处
《机械科学与技术》
CSCD
北大核心
2022年第3期414-420,共7页
Mechanical Science and Technology for Aerospace Engineering
基金
吉林省自然科学基金项目(201554)。
关键词
微纳操纵成像系统
迭代学习
前馈—反馈控制
收敛性分析
micro-nano manipulation imaging system
iterative learning
feedforward and feedback control
convergence analysis
