Skip navigation

DSpace

機構典藏 DSpace 系統致力於保存各式數位資料(如:文字、圖片、PDF)並使其易於取用。

點此認識 DSpace
DSpace logo
English
中文
  • 瀏覽論文
    • 校院系所
    • 出版年
    • 作者
    • 標題
    • 關鍵字
  • 搜尋 TDR
  • 授權 Q&A
    • 我的頁面
    • 接受 E-mail 通知
    • 編輯個人資料
  1. NTU Theses and Dissertations Repository
  2. 電機資訊學院
  3. 電機工程學系
請用此 Handle URI 來引用此文件: http://tdr.lib.ntu.edu.tw/jspui/handle/123456789/70225
標題: ROS-Industrial 6軸機器人適應速度控制應用於複雜曲面物件加工
Adaptive Speed Control of ROS-Industrial 6 DoF Robot for Machining of Complex Curvature Object
作者: Min Zeng
曾旻
指導教授: 羅仁權(Ren C. Luo)
關鍵字: 機器人切削,適應速度控制,安全機制,深度學習,智慧製造,
Robot machining,Adaptive speed control,Safety mechanism,Deep learning,Intelligent manufacturing,
出版年 : 2018
學位: 碩士
摘要: 工業機器人現今已被廣泛應用在工廠中搬運、焊接等運用,但是卻很少用於加工操作,如磨削,去毛刺,拋光,切割,鑽孔或銑削,而目前,工廠中材料的加工工作大部分是由數控機床(CNC)完成。我們的目的是通過開發一台機械手臂能夠取代CNC做一些輕量級的加工,類似低硬度材料的鐉削,對複雜物件進行去毛邊。此外,由於機器人剛性低,在加工硬質材料時為了保護機器人往往會刻意放慢加工速度,這種做法大大降低了加工效率。在一個完整的加工任務中,根據不同的加工負載來自適應調節加工速度是提高效率的途徑。不僅如此,為了同時引入智慧製造的概念。首先,我們利用Android APP進行機器人加工的遠端控制,其次,工業安全也成為我們要考慮的部分。為了避免機器人傷害到人,非接觸式避障被引入這篇論文。
本文介紹了一種基於ROS的機器人機械手的方法和實現,用於相對較輕的加工,並將自適應速度控制系統與安全機制相結合。我們在我們的NTU機器人實驗室開發了一台6軸工業機器人。機器人在機器人操作系統(ROS)上實施並且使用ROS-Industrial。在適應速度控制的部分,我們通過機器人控制器中各關節力矩值建立力矩模型來評估機器人的TCP(刀具中心點)所受額外力矩的大小,當機器人加工速度越快時,所受額外力矩也會越大。因此,系統會根據額外力矩大小,轉換為速度與功率之間的關係,進而自適應調節加工速度。此外,我們還提出了結合振動傳感器的的方法進行加工檢測,通過機器學習中XGBoost的方法建立異常振動檢測模型,用於判斷加工材料之硬度範圍,並且防止異常振動所造成工件損壞或者斷刀之類的情況發生,與力矩模型融合,進一步適應性控制。智慧製造主要在於一:能夠展現通過移動裝置直接控制機器人加工,命令會以類似訂單處理的形式進行排序管理。二:智慧地建立加工的安全環境,我們在文中有對安全進行定義,主要目的是為了避免傷害到誤入機器人加工範圍的人,而對於其他物體的靠近不做任何改變,例如工廠中移動的手臂車。因此在這方面我們使用基於深度學習的電腦視覺辨識方法YOLO來檢測是否有人進入機器人工作區域。在加工的部分,配合使用DASSAULT 3DEXPERIENCE平台進行機器人的加工路徑的軌跡規劃,允許機器人做材料去除的粗加工及在具有複雜曲面的渦輪葉片上的去毛邊動作。
最後,我們通過實驗展示了結果。
Industrial robots have been widely used in factories for handling and welding, but they are rarely used in machining operations such as grinding, deburring, polishing, cutting, drilling or milling. At present, the processing of materials in factories Most of it is done by CNC machine tools. Our goal is to replace the CNC with some lightweight machining by opening a machine phone, similar to the milling of low-hardness materials, to deburr complex objects. In addition, due to the low rigidity of the robot, in order to protect the robot during the processing of hard materials, the processing speed is often deliberately slowed down, which greatly reduces the processing efficiency. In a complete machining task, adaptive adjustment of the machining speed according to different machining loads is a way to improve efficiency. Not only that, but to introduce the concept of smart manufacturing at the same time. First, we use the Android App for remote control of robot processing. Second, industrial security has become a part of our consideration. In order to prevent robots from harming people, non-contact obstacle avoidance was introduced in this paper.
This paper introduces a method and implementation of a ROS-based robot manipulator for relatively light processing and combining adaptive control systems with safety mechanisms. We developed a 6-axis industrial robot in our NTU robotics lab. The robot is implemented on a robot operating system (ROS) and uses ROS-Industrial. In the adaptive control part, we establish the torque model by the joint torque values ​​in the robot controller to evaluate the extra torque received by the robot's TCP (tool center point). When the robot processing speed is faster, the additional torque is also applied. It will be bigger. Therefore, the system converts the relationship between speed and power according to the extra torque, and adaptively adjusts the machining speed. In addition, we also proposed a method combining vibration sensor for machining detection. The XGBoost method in machine learning is used to establish an abnormal vibration detection model for judging the hardness range of the processed material and preventing damage to the workpiece caused by abnormal vibration or breaking the knife. The situation of the class occurs, blended with the torque model, and further adaptive control. Wisdom manufacturing is mainly about one: it can show that the robot processing is directly controlled by the mobile device, and the order will be sorted and managed in the form of similar order processing. Two: Wisdom to establish a safe environment for processing, we have defined the safety in the text, the main purpose is to avoid hurting people who have entered the robot processing range, and do not make any changes to the proximity of other objects, such as mobile manipulator in the factory. So in this regard we use the deep learning-based computer vision recognition method YOLO to detect if someone is entering the robot work area. In the machining part, the DASSAULT 3DEXPERIENCE platform is used to plan the path of the robot's machining path, allowing the robot to perform roughing of material removal and deburring on turbine blades with complex curvature surfaces. Finally, we show the results through experiments. manufacturing.
URI: http://tdr.lib.ntu.edu.tw/jspui/handle/123456789/70225
DOI: 10.6342/NTU201803423
全文授權: 有償授權
顯示於系所單位:電機工程學系

文件中的檔案:
檔案 大小格式 
ntu-107-1.pdf
  目前未授權公開取用
26.04 MBAdobe PDF
顯示文件完整紀錄


系統中的文件,除了特別指名其著作權條款之外,均受到著作權保護,並且保留所有的權利。

社群連結
聯絡資訊
10617臺北市大安區羅斯福路四段1號
No.1 Sec.4, Roosevelt Rd., Taipei, Taiwan, R.O.C. 106
Tel: (02)33662353
Email: ntuetds@ntu.edu.tw
意見箱
相關連結
館藏目錄
國內圖書館整合查詢 MetaCat
臺大學術典藏 NTU Scholars
臺大圖書館數位典藏館
本站聲明
© NTU Library All Rights Reserved