淡江大學機構典藏:Item 987654321/54204
English  |  正體中文  |  简体中文  |  Items with full text/Total items : 62797/95867 (66%)
Visitors : 3748525      Online Users : 461
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: https://tkuir.lib.tku.edu.tw/dspace/handle/987654321/54204


    Title: 模糊行為決策於視覺自主人形機器人之設計
    Other Titles: Fuzzy-based behavior decision design on a vision-based autonomous humanoid robot
    Authors: 黃楷翔;Huang, Kai-Hsiang
    Contributors: 淡江大學電機工程學系博士班
    翁慶昌;Wong, Ching-Chang
    Keywords: 人形機器人;自主移動機器人;視覺機器人;模糊系統;humanoid robot;Autonomous mobile robot;Vision-based robot;fuzzy system
    Date: 2011
    Issue Date: 2011-06-16 22:09:52 (UTC+8)
    Abstract: 本論文應用模糊概念提出一些行為決策,讓視覺自主人形機器人可以完成「FIRA世界盃機器人足球賽:人形機器人組(HuroCup)」之障礙賽跑(Obstacle Run)、罰球(Penalty Kick)、馬拉松(Marathon)與籃球(Basketball)等四項競賽任務。在障礙賽跑競賽中,本論文提出一個模糊行為決策,是以十五條影像掃描線來擷取機器人週遭環境的障礙物資訊,並以機器人與障礙物的相對距離以及機器人與目的方向的相對角度為模糊系統的輸入變數,此模糊系統所架構的行為決策是在每一次的決策中從六個基本移動動作中決定出最合適的行為動作,讓人形機器人可以在不碰觸到任何障礙物的情況下順利穿越障礙物環境。在罰球競賽中,本論文提出以兩個模糊系統用相同的輸入變數各別獲得移動向量的角度與長度,再將其合成一個最合適的行為動作,並以此架構應用於有限狀態機制行為決策系統中追球狀態的逼近球與射門狀態的瞄準球。此外,在射門之前,本論文亦提出另一個模糊系統來調整機器人射門路線的轉向行為,以提高進球率。在馬拉松競賽中,本論文提出一個循線軌跡追蹤的模糊行為決策,透過依據影像畫面中目標線段的結束點位置分成三種狀況以規劃模糊系統,決定出一個最合適的行為動作。在籃球競賽中,機器人所用的模糊行為決策的設計方式如同罰球競賽所用的行為決策。此外,本論文利用倒傳遞類神經網路訓練逆向運動學來推算機器人拿球時所需伺服機的角度,再利用正向運動學進行驗證。從模擬結果與實驗結果的驗證可知,本論文所提出之模糊行為決策確實可以讓視覺自主人形機器人有效的完成障礙賽跑、罰球、馬拉松與籃球等四項競賽任務。
    In this dissertation, some behavior decisions based on the fuzzy concept are proposed for a vision-based autonomous humanoid robot so that it can accomplish four competition events of Obstacle Run, Penalty Kick, Marathon, and Basketball in HuroCup of FIRA (Federation of International Robot-soccer Association). In the event of Obstacle Run, a fuzzy behavior decision is proposed by using fifteen vision scanning lines to retrieve the information of obstacles around the robot. The relative distance between the robot and the obstacle and the relative angle between the robot and the direction of the destination are considered as the inputs of the fuzzy system. One of six basic motions with a highest value is selected to be the next motion in every decision so that the humanoid robot can avoid obstacles successfully and arrive at the terminal area effectively. In the event of Penalty Kick, the same inputs in two fuzzy systems are considered to obtain the angle and the length of the moving vector separately, then they are combined to be the next behavior motion. The structure is implemented on two states of the finite state transition mechanism containing the behavior of approaching ball in the tracing ball state and the behavior of aiming ball in the shooting ball state. Moreover, another fuzzy system used on the turning behavior of adjusting shooting path before shooting ball is also proposed to increase the probability of goal. In the event of Marathon, a fuzzy behavior decision of trajectory tracing by following a visible line is proposed. Three situations are proposed to design fuzzy systems according the terminal point of the visible line in the image frame so that an appropriate behavior motion is decided. In the event of Basketball, the design method of the behavior decision on the Basketball event is similar to that on the Penalty Kick event. Moreover, backpropagation neural network is used to train inverse kinematics to obtain the angle of motors to let the robot can pick up a ping-pong ball and the direct kinematics is used to check it. From the simulations and experiment results, we can see that the proposed fuzzy-based behavior decisions can let the vision-based autonomous humanoid robot effectively accomplish these four events of Obstacle Run, Penalty Kick, Marathon, and Basketball.
    Appears in Collections:[Graduate Institute & Department of Electrical Engineering] Thesis

    Files in This Item:

    File SizeFormat
    index.html0KbHTML224View/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