基于主动视觉定位的多自由度机械手臂控制研究  

A Multi-Degree of Freedom Manipulator Based on Active Vision Positioning

作　　者：蔡剑锋 明兴莹 李杰[1] 王辛月 谭文婷 袁海云 胡诗琴 

机构地区：[1]西南石油大学电气信息学院,四川成都

出　　处：《人工智能与机器人研究》2017年第3期75-90,共16页Artificial Intelligence and Robotics Research

摘　　要：本文介绍的机械臂是一种介于第二代与第三代机械臂之间的机械臂,采用模糊控制器,利用模糊控制不依赖于对象模型、响应速度快和鲁棒性强的优点实现机械臂控制更准确、稳定,实现自动取物,放物等功能。通过具体建立一个主动机器视觉定位系统,适用于工业机器人对零件工位的精确定位。采用基于区域的匹配和形状特征识别相结合的图像处理方法,能够快速准确地得到物体的边界和质心,进行数据识别和计算,再结合机器人运动学原理控制机器人实时运动以消除此误差,满足工业机器人自定位的要求。Mechanical arm introduced in this paper is a mechanical arm between the second generation and the third generation of the manipulator. The advantages of fuzzy control does not depending on the object model, fast response and strong robustness help mechanical arm to achieve controlling more accurately and stably, fetching automatically, putting objects, and other functions. This paper es-tablishes an active machine vision positioning system, which is used for the precision orientation of the parts. Based on matching of area and shape feature recognition with the combination of image processing method, it can rapidly and accurately get the boundary of the object and the center of mass, which is used to identify the data and calculation. Combined with the principle of robot kinematics to control the real-time movement of the robot and eliminate this error, to meet the requirements of industrial robot self-positioning.

关 键 词：模糊控制器 视觉定位系统 实时运动 

分 类 号：TP2[自动化与计算机技术—检测技术与自动化装置]