基于迭代学习控制的原子力显微镜成像

Atomic Force Microscopy Imaging Based on Iterative Learning Control

原子力显微镜(AFM)在成像过程中要求纳米级的定位精度,利用压电陶瓷扫描器能满足要求。该文针对压电陶瓷的非线性及外部环境干扰带来的不利影响,设计一种基于迭代学习算法的AFM扫描成像控制器。通过将水平平面内的扫描运动转换为路径跟踪控制问题,在跟踪过程中对前一迭代周期的误差信息进行非因果学习,保证输出沿迭代轴的快速收敛性,以获得理想的跟踪性能。路径跟踪仿真和实际系统成像实验表明该算法可以有效改善系统非线性和外部环境干扰带来的不利影响,显著提高原子力显微镜的成像质量。

This paper presents a novel controller design based on iterative learning control(ILC)for atomic force microscopy(AFM)nanopositioning.The controller focuses on eliminating the adverse effects brought by nonlinear of piezoelectric actuator and external environmental interference.Specifically,scanning in the horizontal plane of AFM is regarded as a path tracking control problem and the error information of the previous iteration periods is used to modify the control input,to ensure the fast convergence of the output along the iterative axis.The tracking simulation and AFM imaging experimental results are presented and show that the proposed controller can effectively eliminate the adverse effects and significantly improve the imaging accuracy of AFM.