DC 欄位 |
值 |
語言 |
DC.contributor | 電機工程學系 | zh_TW |
DC.creator | 張經民 | zh_TW |
DC.creator | Chin-Ming Chang | en_US |
dc.date.accessioned | 2005-6-16T07:39:07Z | |
dc.date.available | 2005-6-16T07:39:07Z | |
dc.date.issued | 2005 | |
dc.identifier.uri | http://ir.lib.ncu.edu.tw:88/thesis/view_etd.asp?URN=92521078 | |
dc.contributor.department | 電機工程學系 | zh_TW |
DC.description | 國立中央大學 | zh_TW |
DC.description | National Central University | en_US |
dc.description.abstract | 本篇論文提出一種新的環境辨識技術以應用在無人自走車的導引與避障,環境辨識的方法主要以Hopfield類神經網路來記憶學習環境資料,並以改良的Hopfield類神經網路成功的解決記憶資料清洗不完全的問題。車型機器人控制的方面,本論提出以多次的迴轉來節省車型機器人轉彎所需的前方空間,最後以雙層的模糊控制器來完成車型機器人趨近道路中心行駛的控制。
為了驗證本論文所提出的環境辨識方法,吾人分別設計各種自走車可能遭遇的環境,並利用距離探測系統進行環境的辨識,將測距資訊繪製週遭環境圖。而車型機器人的最小迴轉半徑控制論文中除了提供理論上的證明以外,也以實驗結果證明本方法的可行性。
本文利用視覺影像發展一種新式的測距方法,不需要複雜的影像識別演算法,以追求更低的成本與更快的即時運算速度,來取代傳統的超音波或是雷射光的測量工具,未來希望能將此測距系統成功的應用在複雜環境辨識與即時控制實驗中。 | zh_TW |
dc.description.abstract | In this paper, an environment recognition approach is applied in the obstacle avoiding and guidance of an autonomous Car-Like Mobile Robot. The environment recognition is implemented by the Hopfield neural network. We also provide a modified Hopfield network to reduce the output error. For the sake of the control of car-like mobile robot, a turning strategy is proposed to achieve the minimal turning ratio. Finally, a double layer fuzzy controller is designed to keep the robot cruising on the center of the road.
In order to examine the performance of the proposed environment recognition, we create several possible patterns the mobile robot could meet in the real world. We use a novel distance measurement system to recognize the environment, and to plot the surroundings information. The minimal turning ratio control is not only proved in the theory but also verified by experiment of results.
In this thesis, a novel distance measurement method based on the version image is developed. This system could perform the distance measurement without complicate algorithms. We expect that the proposed nonexpensive and faster measurement system would replace the conventional tools, such as the ultrasonic and infrared methods. In the future, we will apply the environment recognition to achieve the real time control of an autonomous robot in the complicate environment. | en_US |
DC.subject | 距離探測系統 | zh_TW |
DC.subject | 環境辨識 | zh_TW |
DC.subject | 霍普菲爾類神經網路 | zh_TW |
DC.subject | 即時控制 | zh_TW |
DC.subject | 車型機器人 | zh_TW |
DC.subject | car-like mobile robot | en_US |
DC.subject | distance sensor system | en_US |
DC.subject | environment recognization | en_US |
DC.subject | hopfield neural network | en_US |
DC.subject | real-time control | en_US |
DC.title | 以新型距離探測系統完成環境辨識和即時車型機器人控制 | zh_TW |
dc.language.iso | zh-TW | zh-TW |
DC.title | Environment Recognization and Real-Time Control on a Car-Like Mobile Robot via a Novel Distance Sensor System | en_US |
DC.type | 博碩士論文 | zh_TW |
DC.type | thesis | en_US |
DC.publisher | National Central University | en_US |