虚拟纳米手操作方法研究
文献类型：学位论文
| 作者 | 王智宇
|
| 学位类别 | 博士 |
| 答辩日期 | 2014-11-27 |
| 授予单位 | 中国科学院沈阳自动化研究所 |
| 授予地点 | 中国科学院沈阳自动化研究所 |
| 导师 | 王志东 ; 董再励 |
| 关键词 | 原子力显微镜微纳操作 纳米手 概率运动模型 迟滞蠕变补偿 |
| 其他题名 | Research on Virutal Nano-hand strategy for AFM based nanomanipulation |
| 学位专业 | 机械电子工程 |
| 中文摘要 | 原子力显微镜（Atomic Force Microscopy AFM）具有极高的观测分辨率和良好的操作可控性，因而不仅可以对物体表面进行纳米尺度的扫描观测，还可以对微纳米物体进行微小力操作。但是，使用原子力显微镜进行微纳米尺度操作存在如下问题：首先，在使用AFM进行操作时，探针作业位置是无法准确预知的。即探针位置存在不确定性。其次，操作过程中缺乏对操作状态的精确反馈信息，难以实现实时修正补偿，即存在操作状态控制的不确定性。再者，目前尚无法给出精确的纳米尺度物体运动学模型和动力学模型，还难以依据模型控制方法实现精确纳米尺度操作，存在着模型控制的不确定性。因而纳米操作在可靠性、精度和效率等方面还存在巨大的挑战。针对纳米尺度操作中存在的目前无法完全消除的不确定性因素，本论文从两个方面来提高纳米操作的可靠性、效率和精度性能指标。一方面，对控制原子力显微镜探针的驱动机构—压电陶瓷（PZT）的非线性驱动特性进行深入研究，除了对主要的静态非线性迟滞进行补偿外，还对动态特性中的蠕变特性进行分析，与已有的迟滞蠕变补偿不同之处在于如下两点：一、提出影响原子力显微镜探针运行精度的拐点蠕变概念并对其建模；二、改变传统PI模型的基函数内核，把拐点蠕变效应嵌入基函数中，使基函数可以同时描述静态特性和动态特性，从而避免使整个模型由于蠕变模型的加入而破坏模型的整体框架。对压电陶瓷迟滞和蠕变的非线性补偿可以保证对探针的定位不确定性在进行操作时保证尽可能的小。另一方面，对目前为止无法完全消除的定位不确定性本身作为研究对象，分析不确定性的初始条件对操作过程和操作结果的影响机理，从分析和控制操作过程和操作结果的不确定性角度来提高操作的可靠性、高效性和精度等性能指标。解决的途径为：一、对初始条件的不确定性进行基于概率的分析，得到其概率分布参数，并依此建立基于概率的纳米颗粒运动模型，通过模型研究纳米颗粒随着原子力显微镜探针的运动其随机位置分布状态的演化规律。二、设计原子力显微镜探针的推动策略—纳米手，设定纳米手结构参数，通过已建立的概率运动模型来分析纳米手结构参数对操作过程和结果的影响；然后，把基于概率的操作过程和结果与人们所关心的可靠性、高效性和高精度指标相关联，创建评价操作性能的代价函数，研究利用代价函数获得优化的纳米手结构参数的方法。 最后的实验利用已经建立的压电陶瓷迟滞蠕变补偿模型对设计后的纳米手策略进行实施，对纳米颗粒进行的操作验证了纳米手策略的可靠性、高效性和精度性能。 |
| 索取号 | TB383/W39/2015 |
| 英文摘要 | Atomic Force Microscopy can get image of the surface in nanoscale resolution, meanwhile，it can manipulate nanoparticles with its sharp probe. However, there are much more challenges in the AFM based nanomanipulation system, as is not faced in macro robotic manipulation systems. The first challenge is that the tip localization uncertainty can not be ignored as it was in the macro world. At the same time, this localization uncertainty will maintain for a considering long time because of both of the actuator nonlinear driving characteristic and its model uncertainty；the second challenge is that the feedback signal can not be easily obtained during the pushing process, since the one single tip can not do the scanning work without stopping the pushing process, to shift between scanning and pushing frequently will slow down the pushing efficiency；the third challenge is that the kinematics and dynamic model in nanoscale are still under studied, while the existing models are approximate analysis. All these challenges should be overcome to earn a higher AFM based nanomanipulation performance with reliability and efficiency, as well as accuracy. The upper mentioned challenges can be concluded as existing uncertainties，which will be considered in this paper. To strain the tip localization uncertainty as small as possible, the paper proposed an extended PI (EPI) model for better AFM tip positioning precision. In the EPI model, a concept of inflexion creep was firstly proposed with its new expression, the parameter of the inflexion creep was introduced in the basic operator of the traditional PI model, therefore the new basic operator can descript both the hysteresis and creep characteristic. The advantage of this modification is maintain the traditional frame of the PI model and restrain the number of introduced parameter as small as possible. The unified hysteresis and creep compensation for AFM tip positioning is the first step for better AFM based nanomaniuplation, the second step is to solve the remained uncertainties as is not completely compensated by the traditional feed forward control strategy. In this paper, the initial uncertainty was described as probabilistic distribution; the kinematics and dynamics of the nanoparticle pushing process were expressed with the firstly proposed probabilistic nanomanipulation motion model; using this model, the stochastic state of nanoparticle pushing could be deduced and described, then the performance of the pushing strategy could be canalized; the cost function was proposed for evaluating the performance of the proposed Virtual Nano-Hand Structure (VNHS), using the cost function, the optimized pushing strategy basing on the VNHS will be adopted. The AFM based nanomanipulation systems precede the optimized VNHS in nanoparticle pushing, which shows the stability, efficiency of the VNHS. |
| 语种 | 中文 |
| 产权排序 | 1 |
| 页码 | 114页 |
| 源URL | [http://ir.sia.ac.cn/handle/173321/16807]  |
| 专题 | 沈阳自动化研究所_机器人学研究室 |
| 推荐引用方式 GB/T 7714 | 王智宇. 虚拟纳米手操作方法研究[D]. 中国科学院沈阳自动化研究所. 中国科学院沈阳自动化研究所. 2014. |
入库方式： OAI收割
来源：沈阳自动化研究所
浏览0
下载0
收藏0
其他版本
除非特别说明，本系统中所有内容都受版权保护，并保留所有权利。