請用此 Handle URI 來引用此文件:
http://tdr.lib.ntu.edu.tw/jspui/handle/123456789/99617| 標題: | 具軟式夾爪與欠驅動手臂之自主移動機器人設計及自主導航應用 Design and Application of an Autonomous Mobile Robot with a Soft Gripper and Underactuated Manipulator |
| 作者: | 陳易宏 Yi-Hung Chen |
| 指導教授: | 郭重顯 Chung-Hsien Kuo |
| 關鍵字: | 自主移動機器人,欠驅動手臂,欠驅動機構控制,軟式夾爪,人工勢能場導航,物件辨識, Autonomous Mobile Robot,Underactuated Manipulator,Control of Underactuated Mechanisms,Soft Gripper,Artificial Potential Field for Navigation,Object Detection, |
| 出版年 : | 2024 |
| 學位: | 碩士 |
| 摘要: | 本研究旨在開發可搭載於自主移動機器人之欠驅動機械手臂,並設計一具有四隻爪子之夾爪作為手臂之末端致動器。本研究使用之自主移動機器人為四舵輪機構,透過運動學的轉換,使其具備差速運動模式以及全向移動模式,並且具備里程計之回授。此外,自主移動機器人配備光學雷達以及慣性測量單元,並透過人工勢能場實現路徑規劃,使機器人能夠建置地圖並且導航及自主避障。經驗證,導航過程中最高時速可達每秒0.67公尺。此外,本研究開發兩種不同的欠驅動手臂,其均為三節串列式連桿機構,且透過拉線驅動使手臂運作,分別採用3D列印以及金屬製造。手臂上的每顆關節均配置絕對式磁編碼器,結合諧波減速機可實現高精度之位置控制,並且混合速度控制以及電流控制架構,實現響應快速且穩定的線驅動控制系統,實驗驗證各關節的平均誤差均小於0.2度,可滿足大多應用場合之需求。最後,本研究開發的軟式夾爪採用3D列印製成,並且透過線性軸承以及簡易四連桿機構使夾爪能以拉線方式驅動。各個爪子均為多材料之機構,使夾爪具有更好的物件輪廓順應性,此外,夾爪搭載多個感測器,結合物件辨識模型,可實現物件辨識、電流控制等功能。實驗證實透過物件辨識模型,夾爪可分辨不同種類的水果,並及時調整爪力進行物件夾取任務。 This study focuses on developing an underactuated manipulator mounted on an autonomous mobile robot (AMR) with a four-fingered gripper as its end effector. The AMR utilizes a four-steering-wheel mechanism to support both differential-drive and omnidirectional motion, providing odometer feedback. It is equipped with a 2D LiDAR and an inertial measurement unit (IMU), utilizing an artificial potential field approach for path planning, navigation, and obstacle avoidance. The robot achieved a maximum navigation speed of 0.67 m/s. Two versions of the underactuated manipulator were designed, consisting of a three-link serial-chain mechanism driven by cables: one made from 3D printing and the other from metal. Each joint features an absolute magnetic encoder and a harmonic reducer for high-precision control, maintaining an average tracking error below 0.2°. A soft, four-fingered gripper was developed using 3D printing and a four-bar linkage for cable-driven actuation. Each finger is designed for improved contour compliance and includes sensors for real-time object detection and adaptive force control. Experimental results confirm the gripper can distinguish various types of fruit and adjust its grasp accordingly. |
| URI: | http://tdr.lib.ntu.edu.tw/jspui/handle/123456789/99617 |
| DOI: | 10.6342/NTU202504199 |
| 全文授權: | 未授權 |
| 電子全文公開日期: | N/A |
| 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-113-2.pdf 未授權公開取用 | 7.25 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。
