English  |  正體中文  |  简体中文  |  全文筆數/總筆數 : 49378/84106 (59%)
造訪人次 : 7383477      線上人數 : 75
RC Version 7.0 © Powered By DSPACE, MIT. Enhanced by NTU Library & TKU Library IR team.
搜尋範圍 查詢小技巧:
  • 您可在西文檢索詞彙前後加上"雙引號",以獲取較精準的檢索結果
  • 若欲以作者姓名搜尋,建議至進階搜尋限定作者欄位,可獲得較完整資料
  • 進階搜尋
    請使用永久網址來引用或連結此文件: http://tkuir.lib.tku.edu.tw:8080/dspace/handle/987654321/35852


    題名: 輪型足球機器人之整合型小腦模型控制器設計與路徑規劃
    其他題名: Integrated CMAC design and path planning for wheeled soccer robots
    作者: 駱佑瑋;Lo, Yu-wei
    貢獻者: 淡江大學電機工程學系碩士班
    翁慶昌;Wong, Ching-chang
    關鍵詞: 模糊控制;小腦模型控制器;足球機器人;運動控制;路徑規劃;Fuzzy Control;CMAC;Soccer robot;Motion control;Path Planning
    日期: 2006
    上傳時間: 2010-01-11 07:14:07 (UTC+8)
    摘要: 本論文提出一個整合型小腦模型控制器來改善輪型機器人之運動控制和運動策略之軌跡控制。此外,在足球機器人之決策判斷上,本論文提出一個進攻策略與軌跡演算法來讓機器人在動態的足球比賽中具備及時選擇進攻路徑之能力。在運動控制上,本論文結合模糊控制器(Fuzzy Controller)與最簡架構之類化型小腦模型控制器(S_CMAC_GBF)提出一個整合型小腦模型控制器(Intergrated CMAC)結構S_FCMAC_GBF來決定輪型機器人逼近目標物時位移速度和旋轉角度之控制量,所提方法將可以控制機器人同時達到位置與角度之雙目標控制要求。S_CMAC_GBF具備快速的學習收斂速度、良好的類化能力、架構簡單、以及易於硬體實現等特性,並且可以大幅減少傳統小腦模型架構所需的記憶體空間。所提S_FCMAC_GBF控制架構在設計過程中不需經過繁雜的系統建模和參數最佳化,其可以經由小腦模型控制器S_CMAC_GBF的線上學習來修正模糊控制器的控制命令來讓機器人達到所需求的位置與角度。所提控制架構對於非線性系統之即時控制具有不錯的強健性和準確度,而且此控制架構具有良好的適應性,所以易於移植到不同的受控系統。在機器人進攻之軌跡控制上,本論文結合比例控制器(Proportional Controller)與最簡架構之類化型小腦模型控制器(S_CMAC_GBF)提出一個整合型小腦模型控制器結構S_PCMAC_GBF來修正機器人之運動軌跡,使得機器人之追球路徑可以逼近追球軌跡演算法之期望路徑。最後,本論文以MATLAB建立一平台來模擬機器人運動軌跡之追蹤控制以及進攻策略與軌跡控制器之運動路徑,並且驗證所提控制架構確實可以有效降低動態運動控制中的控制誤差,以及有效提升系統控制的精確度。
    An integrated CMAC structure is proposed to improve the motion control and the path planning of wheeled robots in this thesis. Moreover, an attacking strategy and a path algorithm are also proposed so that a good attacking path can be real-time generated for soccer robots in the robot soccer game. In the motion control, an integrated CMAC structure S_FCMAC_GBF, combines a fuzzy controller and a S_CMAC_GBF (a simple structure of addressing technique for CMAC_GBF), is proposed to determine the velocity and angular velocity of the robot to approach the target so that the controlled wheeled robot can attain two control targets: the desired position and orientation of the robot. The proposed S_FCMAC_GBF structure does not need complex system modeling or parameter optimal adjusting. S_CMAC_GBF has some distinguished attributes, such as fast learning converge capability, outstanding classifying capability, simple structure, and effectiveness to be implemented in hardware, reducing the memory size of traditional CMAC structure dramatically, etc. Through the on-line learning of S_CMAC_GBF, the output of S_CMAC_GBF can modify the output of the fuzzy control such that the control command will let the controlled robot approach the desired position and orientation as well as possible. The proposed structure has a good robustness and accuracy for the real-time control of nonlinear systems to track the target. Moreover, it is facile to be implemented on different systems. In the attacking path planning design, an integrated CMAC structure S_PCMAC_GBF, combines a proportional controller (P controller) and a S_CMAC_GBF, is proposed to adjust the tracking path so that the robot can move as the expected path determined by the tracking ball algorithm. Finally, some control results of the proposed integrated CMAC structures in the motion control and path planning are simulated by MATLAB to illustrate the proposed structures can reduce the output error and improve the control precision effectively.
    顯示於類別:[電機工程學系暨研究所] 學位論文

    文件中的檔案:

    檔案 大小格式瀏覽次數
    0KbUnknown230檢視/開啟

    在機構典藏中所有的資料項目都受到原著作權保護.

    TAIR相關文章

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