SIA OpenIR  > 机器人学研究室
微装配中微接触力的感知与控制方法研究
Alternative TitleResearch on Sensing and Control of Micro Contact Force in Microassembly
刘意杨1,2
Department机器人学研究室
Thesis Advisor王越超 ; 董再励
ClassificationTP242
Keyword微装配 微接触力 传感器 阻抗控制 蚁群算法
Call NumberTP242/L76/2010
Pages97页
Degree Discipline模式识别与智能系统
Degree Name博士
2010-06-09
Degree Grantor中国科学院沈阳自动化研究所
Place of Conferral沈阳
Abstract微装配系统是微机电系统(MEMS)发展过程中必不可少的关键技术之一,在微小零件的装配、微型光学系统的组装、MEMS传感器的封装等领域发挥着重要作用。但目前微装配技术中还没有可靠的亚微牛顿(sub-μN)分辨率的接触力感知与控制方法。微装配机器人的终端执行器为柔性臂结构,体积小、质量小、结构灵活,模型极为复杂,传统的感知和控制中使用的有限维模型在微装配中不再适用,而且现有的控制方法也难以实现高精度的微装配。因此,需要研究微装配系统中微接触力的感知与控制方法,以获得微装配机器人作业的高精度和高鲁棒性,提高加工作业的可靠性和成功率。在国家自然科学基金课题《基于无穷维系统方法的微装配中力测量与控制研究》的资助下,本文在对国内外文献调研与综述的基础上,针对亚μN微接触力传感器的研发、微装配机器人柔性臂建模以及机器人柔顺控制方法进行了系统的理论研究、仿真分析以及实验研究,具体如下:   (1)通过研究各种敏感元件的特性,选定PVDF材料作为敏感元件设计了1-D和2-D微接触力传感器结构以及信号处理电路,通过把电荷放大电路设计成高阻抗以平衡PVDF自身超高阻抗的方法,避免了接触力不变时电荷的流失,解决了PVDF传感器难以获得静态信号的问题。在传感器标定环节,由于目前没有亚μN微接触力传感器的标定方法,本文提出了一种新颖的间接标定方法,并给出了标定实验的设计。标定实验的结果证明本文设计、开发的微接触力传感器线性度良好,而且分辨率达到了0.1μN。   (2)在微装配机器人末端执行器建模方面,考虑到PVDF薄膜的弹性在微/纳米尺寸环境中的影响无法忽略,传统建模中的舍项模型在高精度微装配中不再适用,因此在Hamilton原理的基础上,建立出微装配机器人末端执行器的高精度模型以及传感器所受的微接触力与输出电压的精确函数关系;   (3)深入研究了机器人柔顺控制方法的分类以及研究现状,利用已经建立的微装配机器人柔性臂的高精度模型,基于阻抗控制算法设计了微装配机器人的柔顺控制器;   (4)为了完成微接触力传感器的标定,验证柔顺控制器的性能以及完成微装配和微操作的相关实验,搭建了实验平台。并设计了3个实验测试了微接触力传感器以及阻抗控制系统的性能。实验结果表明,微接触力传感器和控制系统的精度都达到了亚μN,证明了传感器与控制器的有效性;   (5)为了提升现有的柔顺控制方法性能,本文提出了一种新颖的基于蚁群模糊神经网络的阻抗控制方法。解决了传统模糊神经网络阻抗控制方法中代表隶属函数参数的权值在学习过程中可能陷入局部极值的问题。并通过仿真研究,分别在正弦、方波以及随机参考输入的情况下,对比现有的模糊神经网络阻抗控制方法的跟踪性能,仿真结果验证了新颖的基于蚁群模糊神经网络阻抗控制方法的优越性。
Other AbstractMicro assembly system is one of the most important key technology in Micro Electro Mechanical System (MEMS). It is very important in the assembly of micro devices, optical systems and the encapsulation of MEMS sensors. But there are no reliable micro contact force sensors in the range of sub-micro Newton (sub-μN) and the corresponding control methods. With complicated model, the end manipulators of micro assembly robots are small, light, and flexible so that traditional finite dimension modeling method is not suitable to the end manipulators. Furthermore, the present control methods for flexible manipulator can not satisfy the micro assembly. Therefore, the micro contact force sensing and control methods are urgently needed to achieve the high precision and robustness of micro assembly robots and to promote the efficiency and reliability of micro manufacture. Based on the project supported by National Natural Foundation of China, “Research on Force Sensing and Control in Micro Assembly According to Infinite Dimension System Method”, according to lots of research, the micro contact force sensor, flexible manipulator, and the flexible control method are designed and realized in this paper. The outlines are shown as below:  (1) According to the features of many kinds of sensitive material, the Polyvinylidene floride (PVDF) is used to design the 1-D and 2-D micro contact force sensors. The corresponding signal processing circuit is designed and fabricated. By designing the charge amplifier as high impedance to balance the high impedance of PVDF to avoid the charge from disappearing when the force acting on the sensor does not change, the problem that PVDF sensor is difficult to be used to measure static force is resolved. As to the calibration of the sensors, a new indirect calibration method and corresponding experiments are designed and realized in this paper. The results show that the linearity is good, and the resolution is 0.1μN. (2) Because the traditional finite dimensions modeling method can not satisfy the high precision of micro assembly and the flexibility of PVDF film can not be ignored in the scale of micro/nano meter, the accurate model of flexible end manipulator and the relationship between the micro contact force and the output voltage of the sensor are setup based on the Hamilton Principle. (3) Based on the literature review of the flexible control, the impedance control method is used to design the control system of the micro assembly robot’s end manipulator. (4) To complete the calibration, micro assembly experiments and the micromanipulation, to verify the performance of the flexible control system, the experimental platform is setup. Four experiments are designed to test the performance of the 2-D sensor and to verify the performance of impedance control system. The results show that the control precision and sensing resolution is in range of sub-μN.  (5) To promote the performance of present flexible control methods, a novel ant colony optimization- fuzzy neural networks (ACO-FNN) impedance control method is presented, designed and realized. Present FNN impedance control method can not avoid the weights of the FNN from being trapped in local optimal values. But the ACO-FNN is able to solve the above problem. In the simulation, the sine function, square function and random function are respectively used as the reference input of the control system. The experimental results are compared to the FNN impedance control system with the same reference input. The results show that the performance of ACO-FNN impedance control system is better.  In conclusion, the research in this paper provides a feasible method for micro manufacture and micro assembly, and will promote the technology in automatic manufacture and assembly of micro devices.
Language中文
Contribution Rank1
Document Type学位论文
Identifierhttp://ir.sia.cn/handle/173321/9407
Collection机器人学研究室
Affiliation1.中国科学院沈阳自动化研究所
2.中国科学院研究生院
Recommended Citation
GB/T 7714
刘意杨. 微装配中微接触力的感知与控制方法研究[D]. 沈阳. 中国科学院沈阳自动化研究所,2010.
Files in This Item:
File Name/Size DocType Version Access License
微装配中微接触力的感知与控制方法研究.p(2147KB) 开放获取LicenseApplication Full Text
Related Services
Recommend this item
Bookmark
Usage statistics
Export to Endnote
Google Scholar
Similar articles in Google Scholar
[刘意杨]'s Articles
Baidu academic
Similar articles in Baidu academic
[刘意杨]'s Articles
Bing Scholar
Similar articles in Bing Scholar
[刘意杨]'s Articles
Terms of Use
No data!
Social Bookmark/Share
All comments (0)
No comment.
 

Items in the repository are protected by copyright, with all rights reserved, unless otherwise indicated.