智慧型水下聲學載具應用研究

Abstract

台灣政府近年來亟欲發展風力發電，但台灣地狹人稠，陸上風力發電發展不易，因而使海上離岸風力發電為發展項目。離岸風力發電潛力場址與中華白海豚的重要棲息地相近，故相關單位須嚴格控制施工打樁所產生的噪音以避免噪音過大影響中華白海豚，而傳統的監測噪音的方式需要過多的人力與時間成本，為提升監測效率與降低成本，自動化監測方式相應而生。 本文開發出用於監測離岸風機打樁噪音的水上無人載具，將現有的RC遙控無人載具改良成自動化無人載具。將美國麻省理工學院所開發出的開源軟體MOOS-IvP安裝於樹莓派中，利用樹莓派作為無人載具的任務控制電腦與現有的RC遙控無人載具結合，再額外搭載錄音介面與水下麥克風製作成可監測水下噪音的自動化無人載具。本研究所使用的程式語言為C++，運用MOOS-IvP的點對點通訊系統架構、分佈式處理框架以及簡易明瞭的程式撰寫方式開發出適用於MOOS-IvP系統的打樁噪音量測應用程式。將軟硬體相互結合，並在基隆河域上做模擬監測打樁噪音的測試來應證整個智慧型水下聲學載具系統的可行性。

In recent years, the Taiwan government is eager to develop wind power. Cause of the dense population and limited land, Offshore Wind Farms becomes the primary focus. However, the locations of potential offshore wind farms are near the habitat of the Sousa chinensis so the relative units need to control the noise of pile driving of the offshore wind turbine to avoid the noise influences the Sousa chinensis directly. The method of monitoring underwater noise in the past is a waste of time and manpower therefore automated monitor become an important topic. The main purpose of this article is developing an unmanned surface vehicle (USV) for monitoring the noise of pile driving automatically by improving a RC controlled unmanned surface vehicle to an automated unmanned surface vehicle. Use Raspberry Pi to be a mission controlled computer on USV by installing MOOS-IvP developed by Laboratory for Autonomous Marine Sensing System, Massachusetts Institute of Technology. The main programming of this research is C++. According to the advantage of MOOS-IvP such as peer to peer communication system architecture, distributed processing framework and clear programming style, the application for monitoring underwater noise was developed to work in MOOS-IvP system. After combining the all things, USV was tested on the Keelung river in Taiwan for simulation experiment of monitoring the noise of pile driving.