English  |  正體中文  |  简体中文  |  Items with full text/Total items : 55523/89843 (62%)
Visitors : 10960677      Online Users : 191
RC Version 7.0 © Powered By DSPACE, MIT. Enhanced by NTU Library & TKU Library IR team.
Scope Tips:
  • please add "double quotation mark" for query phrases to get precise results
  • please goto advance search for comprehansive author search
  • Adv. Search
    HomeLoginUploadHelpAboutAdminister Goto mobile version
    Please use this identifier to cite or link to this item: http://tkuir.lib.tku.edu.tw:8080/dspace/handle/987654321/35634


    Title: 視覺自主人形機器人之設計與實現
    Other Titles: Vision-based autonomous humanoid robot design
    Authors: 黃楷翔;Huang, Kai-hsiang
    Contributors: 淡江大學電機工程學系碩士班
    翁慶昌;Wong, Ching-chang
    Keywords: 二足機器人;人形機器人;自主機器人;Biped robot;Humanoid robot,;Autonomous robot
    Date: 2006
    Issue Date: 2010-01-11 06:54:05 (UTC+8)
    Abstract: 本論文主要是設計與實現一個多功能且具有高效能的視覺自主人形
    機器人系統。本論文先以繪圖軟體SoildWorks 設計機器人的機構,再以Nios II 發展板架構機器人的運動控制核心,並以BCB 架構一個可以用於設計與規劃機器人動作的人機介面,最後以凌陽科技的u’nsp 十六位元處理器搭配CMOS 感測器架構機器人的自主視覺與策略系統,另外也應用感測器加強機器人對外在環境的適應能力。此人形機器人的架構可分成下列五項:(1)機構部分:機器人的關節由26 顆伺服馬達所構成,並依據各關節施力大小的不同選用不同扭力的伺服馬達,使機器人結構足以實現基本步行與自主射門等動作。(2)運動控制核心部分:機器人以Nios II 發展板做為運動控制核心,經過解碼與分析將機器人的動作資訊儲存至Flash中,待執行動作時才予以讀出以致能伺服馬達動作。(3)人機介面部分:人機介面透過RS232 串列傳輸將所設計完成的動作資料傳送至機器人端,此做法除了能提升動作調整的效率,更可事先透過3D 的動作模擬與馬達角度範圍的鎖定來減少因錯誤動作而導致伺服馬達損傷。(4)影像與策略部分:機器人利用CMOS 感測器擷取影像資訊,以顏色來辨識所要追蹤之物體,並將顏色資訊交由策略端做判斷。在自主射門的策略中,以有限狀態機制為基本架構,分成找球、追球與射門三大狀態,再依據限制條件做策略上的選擇。(5)感測器應用部分:透過陀螺儀、紅外線感測器與電子指南針的應用,讓機器人能克服不平坦的地形並進行障礙物閃避,同時也可以在偏離行進方向時適時做出修正的動作。從基本步行、避障與自主射門等實驗結果的呈現可以證明本論文所設計實現的視覺自主人形機器人的確符合多功能與高效能的要求。
    In this thesis, a design method of a vision-based autonomous humanoid robot system with multiple functions and high efficiency is proposed. The drawing software, SoildWorks, is applied to design the mechanism of robot. A Nios II developmental board is used to be the motion controller of robot. BCB is applied to establish a human-machine interface for designing and planning actions of robot. A 16-bit processor, μ’nsp, made by Sunplus is used to process the image captured by a CMOS sensor to find some objects and make decision. The structure of robot can be described by five items:
    (1) Mechanism: Twenty-six joints are design for the humanoid robot so that it can walk and shoot the ball. Each joint is implemented by a server motor and three types of server motors with different torques are used for this robot according to the desired force of joint.
    (2) Motion control center: A Nios II developmental board is used as a motion control center. Some motion controlled data are constructed by a trial-and-error method and stored in the flash memory. The stored data will be accessed to enable each server motor so that the robot can do the desired motion.
    (3) Human-machine interface: A human-machine interface, which can transmit action data to the robot by a RS-232 serial transmission, is constructed so that the action can be efficiently adjusted by a 3D action simulation. Moreover, the broken probability of server motor can be
    reduced by locking the rotational range of sever motor.
    (4) Vision and tactic: The vision data is captured by a CMOS sensor and one assigned color is applied to recognize the object. A finite-state transition mechanism with three states: (a) find the ball, (b) track the ball, and (c) shoot the ball, is constructed in the tactic so that the
    implemented robot can autonomous shoot. The tactic determines which state is used according to some given condition.
    (5) Sensor application: The gyroscope, infrared sensor, and electronic compass are respectively used to let the robot can walk on an uneven terrain, avoid obstacles, and adjust the walk direction by itself.

    Some experimental results of basic walk, obstacle avoidance, and autonomous shoot are presented to illustrate that the implemented vision-base autonomous humanoid robot meets the requirements of multiple function and high efficiency.
    Appears in Collections:[Graduate Institute & Department of Electrical Engineering] Thesis

    Files in This Item:

    File SizeFormat
    0KbUnknown279View/Open

    All items in 機構典藏 are protected by copyright, with all rights reserved.


    DSpace Software Copyright © 2002-2004  MIT &  Hewlett-Packard  /   Enhanced by   NTU Library & TKU Library IR teams. Copyright ©   - Feedback