請用此 Handle URI 來引用此文件:
http://tdr.lib.ntu.edu.tw/jspui/handle/123456789/90570
標題: | 以高精度雷射線掃描校正機械手臂之精度 Precision Calibration Technique for a Manipulator Using a High-Accuracy Laser Scanner |
作者: | 劉佳鑫 Jia-Xin Liu |
指導教授: | 林沛群 Pei-Chun Lin |
關鍵字: | 機械手臂,線掃描,精度校正, Robotic arm,Laser scanner,Accuracy calibration, |
出版年 : | 2023 |
學位: | 碩士 |
摘要: | 本文討論了商用型機械手臂的精度問題,指出其在點精度和軌跡精度方面的挑戰性。相比CNC工具機的精度,機械手臂的精度通常較低,這導致在需要高精度的任務中,如焊接、3D列印、點膠、研磨和拋光等,人工教導軌跡點成為一種過度依賴的方法,從而降低了生產效率。
為了提高機械手臂的精度,傳統校正手臂精度是使用雷射追蹤儀,但其價格昂貴、設備架設彈性有限,故本研究提出了一個使用雷射線掃來進行精度優化,來校正精度的流程,該流程包括以下步驟: 1.刀具座標系校正:建立一個準確的刀具座標系,以便後續的量測和校正工作。 2.開發工件座標系校正:設計量具並分析量具尺寸,以減小量測誤差對計算誤差的影響,從而獲得更準確的工件座標系。 3.靜態精度校正:推導線性化誤差模型和非線性化誤差模型,使用最小方差法和最佳化技術計算DH參數的誤差,從而降低靜態精度。 4.動態軌跡精度校正:使用線性脈衝響應模型來減少位置誤差,從數據結果來看,目前的效果有改善空間,提空了後續研究和改進的參考依據。 透過這些校正方法,可以提高機械手臂的精度,從而確保其在不同生產製令單中能夠保持良好的製造精度並減少人力需求。然而,對於動態軌跡精度的校正仍存在挑戰,需要更深入的研究和改進。 此外也提到了雷射追蹤儀作為量測機械手臂誤差的首選儀器,但由於其價格昂貴、設備架設彈性有限,通常只在機械手臂量測後使用。因此考慮到生產過程中的量測需求,研究採用雷射線掃描作為替代方案,以獲得更準確的量測結果。 本研究提供了商用型機械手臂精度校正的流程和方法,旨在改善其點精度和軌跡精度。這將對提高機械手臂在高精度任務中的應用性和生產效率具有重要意義,然而仍需要進一步的研究和改進,特別是在動態軌跡精度的校正方面。 This paper discusses the accuracy issues of commercial robotic arms, highlighting the challenges they face in terms of point precision and trajectory precision. Compared to CNC machine tools, robotic arms generally exhibit lower accuracy, which leads to an overreliance on manual teaching of trajectory points in tasks that require high precision, such as welding, 3D printing, dispensing, grinding, and polishing, thereby reducing production efficiency. To improve the accuracy of robotic arms, the traditional approach involves using laser trackers for calibration. However, due to their high cost and limited flexibility in equipment setup, this study proposes a calibration process using laser line scanning to optimize accuracy. The process includes the following steps: 1.Tool coordinate system calibration: Establishing an accurate tool coordinate system to facilitate subsequent measurements and calibration work. 2.Development of workpiece coordinate system calibration: Designing gauges and analyzing their dimensions to minimize the influence of measurement errors on calculation errors, thereby obtaining a more accurate workpiece coordinate system. 3.Calibration of static precision: Deriving linearized error models and non-linear error models, utilizing the least squares method and optimization techniques to calculate the errors of Denavit-Hartenberg (DH) parameters, thereby reducing static precision. 4.Calibration of dynamic trajectory precision: Using linear pulse response models to reduce position errors. Although the results did not meet the expectations, they provide a reference for subsequent research and improvement. By employing these calibration methods, the precision of robotic arms can be enhanced, ensuring consistent manufacturing accuracy and reducing the need for manual labor across different production orders. However, challenges remain in calibrating dynamic trajectory precision, requiring further research and improvement. Furthermore, the study mentions that laser trackers are preferred instruments for measuring robotic arm errors. However, their high cost and limited flexibility in equipment setup result in their usual application only after robotic arm measurements are completed. Therefore, considering the measurement needs during the production process, laser line scanning is adopted as an alternative solution to achieve more accurate measurements. This study provides a process and methods for calibrating the accuracy of commercial robotic arms, with the aim of improving their point precision and trajectory precision. This will have significant implications for enhancing the applicability and production efficiency of robotic arms in high-precision tasks. Nevertheless, further research and improvement are required, particularly in the calibration of dynamic trajectory precision. |
URI: | http://tdr.lib.ntu.edu.tw/jspui/handle/123456789/90570 |
DOI: | 10.6342/NTU202303404 |
全文授權: | 未授權 |
顯示於系所單位: | 機械工程學系 |
文件中的檔案:
檔案 | 大小 | 格式 | |
---|---|---|---|
ntu-111-2.pdf 目前未授權公開取用 | 6.59 MB | Adobe PDF |
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。