基于无穷维系统方法的微装配机器人三维微力信息感知与智能控制方法研究

项目名称： 基于无穷维系统方法的微装配机器人三维微力信息感知与智能控制方法研究

项目编号： No.61203329

项目类型： 青年科学基金项目

立项/批准年度： 2013

项目学科： 自动化学科

项目作者： 刘意杨

作者单位： 中国科学院沈阳自动化研究所

项目金额： 25万元

中文摘要： 自动化微装配是微机电系统（MEMS）必不可少的关键技术之一。但目前微装配中还没有可靠的亚微牛顿分辨率的三维微接触力感知与控制方法，无法满足微镜片等易碎微设备对于高精度自动化装配的迫切要求。本课题针对上述问题，在自主研发的一维、二维高灵敏度亚微牛顿分辨率传感器的基础上，进一步研究微装配机器人的三维微接触力感知、建模与控制方法。首先设计三维微接触力传感器来获取实时、可靠的微接触力信息；其次针对微装配机器人末端执行器的柔性臂结构，基于无穷维系统方法，建立精确模型；最后利用该模型，研究基于蚁群模糊神经网络的智能阻抗控制方法，实现微装配机器人的高精度操作。本课题研究成果为微尺度下的自动化操作与装配提供了一种可行的方法，将提高微系统高精度装配制造的自动化水平，对于实现微设备的可靠、高精度的批量制造具有重要意义。

中文关键词： 微装配机器人；三维微接触力感知；无穷维系统方法；蚁群模糊神经网络；智能阻抗控制

英文摘要： Automatic micro-assembly is pivotal technology necessary for Micro Electromechanical Systems(MEMS). However, at present, in micro-assembly, there is no reliable methods to sense and control three dimensions micro contact force in the range of sub-micro-newton(sub-μN), which can not satisfy that fragile micro devices such as micro mirrors are automatically assembled with high precision. To solve the above problem, based on the 1D and 2D micro contact force sensors with resolution of sub-μN developed by us, a novel 3D micro contact force sensor will be designed, the complex model of micro-assembly robot will be setup, and the intelligent control method will be carried out in this research. Firstly, a 3D micro contact force sensor is designed to sense the reliable and real time micro contact force; secondly, as to the flexible arm of micro-assembly robot's end manipulator, the model is setup based on infinite dimensions method; last but not least, with the above complex model, a novel intelligent impedance control method founded on Ant Colony Optimization and Fuzzy Neural Network is researched to make the high precision manipulation of micro-assembly robots come true. This research will provide a feasible method for automatic micro-assembly and micro-manipulation, will improve the complexity and automation of high

英文关键词： micro-assembly robots；3D micro contact force sensing；infinite dimensions system method；ant colony optimization based fuzzy neural network；intelligent impedance control