整合視覺、力量感測與模糊控制技術之新穎機械手臂組裝系統;A Novel Robotic Assembly System Integrating Vision, Force Sensing, and Fuzzy Control Technologies

NCU Institutional Repository > 工學院 > 光機電工程研究所 > 博碩士論文 > Item 987654321/96150

jsp.display-item.identifier=請使用永久網址來引用或連結此文件: http://ir.lib.ncu.edu.tw/handle/987654321/96150

题名:	整合視覺、力量感測與模糊控制技術之新穎機械手臂組裝系統;A Novel Robotic Assembly System Integrating Vision, Force Sensing, and Fuzzy Control Technologies
作者:	李家銘;Li, Jia-Ming
贡献者:	光機電工程研究所
关键词:	模糊控制;力量控制;機器視覺;機械手臂組裝;伺服運動控制;Robot Operating System (ROS);Fuzzy Control;Force Control;Machine Vision;Robotic Assembly;Servo Motion Control;Robot Operating System (ROS)
日期:	2024-11-25
上传时间:	2025-04-09 16:14:27 (UTC+8)
出版者:	國立中央大學
摘要:	本研究在Linux系統的Ubuntu 20.04版本上，利用機器人作業系統(Robot Operating System, ROS)開發控制軟體。透過ROS的點對點網路和分散式架構，整合了所有的資訊，並實現了工業電腦(Industrial PC)、六軸工業機械手臂、立體相機、六軸力量/力矩感測器以及自適應夾爪等硬體設備與軟體系統的協同運作。本研究針對機械手臂在齒輪箱組裝任務中的挑戰，提出了一種結合視覺與力量控制的模糊控制系統。傳統的順應性控制方法往往需要手動調整多個參數，以應對手臂初始校正、視覺定位誤差以及每次零件夾取時的變異，這導致了自動化組裝過程繁瑣且效率不高。為了解決這些問題，本研究設計並實現了一套多輸入多輸出(Multiple-Input Multiple-Output, MIMO)的模糊力量控制系統，使機械手臂能夠根據即時的組裝情況、力量反饋和視覺定位等感測資訊進行動態調整控制參數，從而提高手臂控制系統的自適應性。經過一系列的實驗驗證結果顯示，所提出之新系統在機車齒輪箱組裝過程，相較於先前本研究室所開發之前系統，減少34%的組裝時間，平均降低30%的最大組裝力量及27%的組裝力量標準差，顯著提升組裝效率和組裝穩定度。此外，也再以行星齒輪系統進行組裝驗證實驗，再次證明了模糊控制在此類複雜干涉與定位的自動化齒輪系統組裝任務中的創新性和實務應用性。 ;This research developed control software on the Ubuntu 20.04 version of the Linux system, using the Robot Operating System (ROS). Through ROS′s peer-to-peer network and distributed architecture, all information was integrated, facilitating the coordinated operation of industrial PC, a six-axis industrial robot arm, a stereo camera, a six-axis force/torque sensor, and adaptive grippers. Addressing the challenges in gearbox assembly tasks for robotic arms, this research proposed a fuzzy control system that combined vision and force control. Traditional compliance control methods often required manual adjustment of multiple parameters to accommodate initial arm calibration, visual positioning errors, and variations in each part-grabbing process, leading to a tedious and inefficient automated assembly process. To overcome these limitations, this study designed and implemented a Multiple-Input Multiple-Output (MIMO) fuzzy control system that dynamically adjusted control parameters based on real-time assembly conditions, force feedback, and visual positioning sensor data, thus enhancing the adaptability of the robotic arm control system. A series of experimental results demonstrated that the proposed new system significantly improved assembly efficiency and stability in motorcycle gearbox assembly tasks compared to the previously developed system, showing a 34% reduction in assembly time, an average 30% decrease in maximum assembly force, and a 27% reduction in force standard deviation. Furthermore, additional validation through planetary gear system assembly experiments confirmed the innovative and practical application of fuzzy control in complex, interference-sensitive, and precision-demanding automated gear system assembly tasks.
显示于类别:	[光機電工程研究所 ] 博碩士論文

文件中的档案:

档案	描述	大小	格式	浏览次数
index.html		0Kb	HTML	10	检视/开启

在NCUIR中所有的数据项都受到原著作权保护.

社群 sharing

数据加载中.....