請用此 Handle URI 來引用此文件:
http://tdr.lib.ntu.edu.tw/jspui/handle/123456789/22134
標題: | 海洋探測用自主式水下載具系統設計與驗證 System Design and Verification of an Autonomous Underwater Vehicle for Ocean Survey |
作者: | Sheng-wei Huang 黄盛煒 |
指導教授: | 郭振華(Jen-hwa Guo) |
關鍵字: | 自主式水下載具,流線扁平型載具,載具能量估測,單聲標定位系統,側掃聲納掃描線海床分類, autonomous underwater vehicle,flat type vehicle,vehicle energy consumption estimation,single beacon positioning system,seafloor classification by sidescan sonar scanline, |
出版年 : | 2018 |
學位: | 博士 |
摘要: | 自主式水下無人載具常應用於海洋探測任務之中,如水下海事工程或軍事除雷作業等高風險的任務,本論文根據實際應用需求建立了一套智慧型自主式無人水下載具系統,包含了扁平流線外型、載具耗能預估、單點式聲標定位與流速量測、載具探測空間和時間的誤差。流線外型使得載具具備良好的操控性能和穩定性能,以符合大多數應用的需求。利用載具外型阻力可以估算載具上有限的能源下探測固定一個空間時的空間探測誤差和時間探測誤差,本論文中的載具在不考慮外部水流條件下,最佳探測速度1.26m/s,在2.88Kw-hr的電池容量下,載具可探測路徑長度為86.5公里(約46.71海浬)。載具的定位也對探測任務十分重要,在本論文中使用單點聲標來協助載具的位置推測,藉由具備高精準度的同步時鐘的海洋聲層析設備,在台灣近岸的海灣中進行驗證,確實可以有效的降低載具的定位誤差,定位誤差可以由單純慣性導航的80公尺下降至27.58公尺以內,足以提供再次搜索的定位參考;利用載具上的側掃聲納的掃描線,透過擷取掃描線中的訊號能量、分布參數、分布誤差三種特徵,可以有效的分類海底的底床,由近岸測試結果顯示本論文提出的方法對於泥和沙底質等軟質的海床辨識率可達91.37%,驗證本論文之成果可以應用在實務之中。 Autonomous underwater vehicle often used in high-risk tasks such as underwater construction work or military mine hunting. This work builds a flat type autonomous underwater vehicle for this kind of application. This work consists of flat type appearance design, vehicle energy analysis, single-beacon positioning system and target identification system. The flat streamline let vehicle not only had good maneuverability but also had good stability to meets most of applications requirement. The survey space error and temporal error can be identified in limited energy on the vehicle. In this paper, the optimum survey speed is 1.26m/s without considering the current in the field and survey ability is 86.5 kilometers (46.71 nautical mile) long with battery capacity of 2.88Kw-hr. The localization is another key issue in survey job. The AUV is localized by dead-reckoning and measure one way travel time from a beacon. The ocean acoustic tomography beacon with high precision synchronous clock is used in this work to verify in the bay area. The positioning error is reduced from 80 meter to 27.58 meter and good enough for next surveyor reference.. Three features of sidescan signal, which are signal energy, distribution shape parameter and distribution error are used to classify the seafloor. The mud class and sand class recognition rate reach to 91.37%. The results of this work show the feasibility of AUV in maritime engineering and military filed. |
URI: | http://tdr.lib.ntu.edu.tw/jspui/handle/123456789/22134 |
DOI: | 10.6342/NTU201802312 |
全文授權: | 未授權 |
顯示於系所單位: | 工程科學及海洋工程學系 |
文件中的檔案:
檔案 | 大小 | 格式 | |
---|---|---|---|
ntu-107-1.pdf 目前未授權公開取用 | 10.2 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。