冯朝, 凌杰, 明敏, 肖晓晖. 融合迭代学习与干扰观测器的压电微动平台精密运动控制[J]. 机器人, 2018, 40(6): 825-834. DOI: 10.13973/j.cnki.robot.170492
引用本文: 冯朝, 凌杰, 明敏, 肖晓晖. 融合迭代学习与干扰观测器的压电微动平台精密运动控制[J]. 机器人, 2018, 40(6): 825-834. DOI: 10.13973/j.cnki.robot.170492
FENG Zhao, LING Jie, MING Min, XIAO Xiaohui. Precision Motion Control for a Piezoelectric Micro-positioning Stage via Integrating Iterative Learning and Disturbance Observer[J]. ROBOT, 2018, 40(6): 825-834. DOI: 10.13973/j.cnki.robot.170492
Citation: FENG Zhao, LING Jie, MING Min, XIAO Xiaohui. Precision Motion Control for a Piezoelectric Micro-positioning Stage via Integrating Iterative Learning and Disturbance Observer[J]. ROBOT, 2018, 40(6): 825-834. DOI: 10.13973/j.cnki.robot.170492

融合迭代学习与干扰观测器的压电微动平台精密运动控制

Precision Motion Control for a Piezoelectric Micro-positioning Stage via Integrating Iterative Learning and Disturbance Observer

  • 摘要: 针对运动系统中常见的重复参考轨迹,尽管迭代学习控制(iterative learning control,ILC)可以通过迭代有效消除重复误差,但其对于非重复性干扰十分敏感.为实现在非重复干扰环境下压电微动平台的精密运动,提出了融合ILC与干扰观测器(disturbance observer,DOB)的控制策略.为避免复杂的迟滞建模,将迟滞非线性视为迭代过程中的重复性输入干扰.为保证控制策略的稳定性,推导其收敛条件并分析对非重复性干扰的抑制作用从而降低收敛误差.最后在压电微动平台进行了对比实验,结果表明:所提控制策略可以在无迟滞模型的前提下有效补偿迟滞非线性.针对理想环境下的5Hz、10Hz、20Hz三角波跟踪,其跟踪误差的均方根在行程的0.4%以内;而在非重复干扰环境下,跟踪误差的均方根为10.24nm,与内置的控制器、单独的反馈控制器、ILC相比,分别降低了98.73%、98.67%与88.24%.而且在干扰环境下,所提控制策略加快了ILC的收敛速度.实验结果充分验证了所提控制策略的有效性,实现了压电微动平台的精密运动.

     

    Abstract: For the commonly used repetitive trajectory in motion systems, the iterative learning control (ILC) method is still quite sensitive to non-repetitive disturbance, although it can eliminate repetitive errors effectively via iterations. In order to achieve precision motion with non-repetitive disturbance for piezoelectric micro-positioning stages, a control strategy integrating ILC with disturbance observer (DOB) is proposed. Firstly, the hysteresis nonlinearity is treated as repetitive input disturbance during the iterative process to avoid complex hysteresis modeling. Then, to ensure the stability of the proposed strategy, the convergence condition is deduced and the suppression of non-repetitive disturbance is analyzed to minimize convergence error. Finally, comparative experiments are performed on a piezoelectric micro-positioning stage. Results show that the proposed strategy can compensate hysteresis nonlinearity effectively without hysteresis modeling. The root-mean-square values of tracking errors are within 0.4% of the stroke for tracking of 5Hz,10Hz and 20Hz triangular waves under ideal environment. While for the environment with non-repetitive disturbance, the proposed strategy can achieve root-mean-square value of tracking error at 10.24nm, that is reduced by 98.73%, 98.67% and 88.24% respectively compared with the built-in controller, the stand-alone feedback controller and ILC. Besides, the proposed control strategy can accelerate the convergence speed of ILC. Experimental results validate the effectiveness of the proposed strategy sufficiently, and the precision motion of the piezoelectric micro-positioning stage can be realized.

     

/

返回文章
返回