複合式類神經網路預測貨櫃船主機油耗;Prediction of Main Engine Fuel Consumption of Container Ships Based on Hybrid Artificial Neural Network

NCU Institutional Repository > 工學院 > 機械工程學系碩士在職專班 > 博碩士論文 > Item 987654321/90113

jsp.display-item.identifier=請使用永久網址來引用或連結此文件: http://ir.lib.ncu.edu.tw/handle/987654321/90113

题名:	複合式類神經網路預測貨櫃船主機油耗;Prediction of Main Engine Fuel Consumption of Container Ships Based on Hybrid Artificial Neural Network
作者:	傅國唐;Fu, Kuo-Tang
贡献者:	機械工程學系在職專班
关键词:	氣候變遷;碳排放;複合式類神經網路;Global Warming;Carbon Emission;hybrid artificial neural network
日期:	2022-07-11
上传时间:	2022-10-04 12:11:27 (UTC+8)
出版者:	國立中央大學
摘要:	2023年起，國際航線船舶必須符合國際海事組織2023年EEXI（現有船舶能效指標）和CII（碳強度指標）要求，這些要求旨在大幅降低碳排放總量和碳排放強度。海事組織進一步制定戰略，要求船舶在2030年與2008年相比要減少40%的碳排放強度，到2050年更進一步要減少70%。因此，船舶運營商目前正在尋求可能的解決方案，以實現這一排放目標。在2030年之前，大多數船隻可通過優化船體設計和船舶速度調整（操作）方案進行碳排放強度的改善。而船速調整必須仰賴主機在各種船速與運行條件下的精準油耗預測結果，本研究是採用複合式類神經網路來預測主機油耗，與傳統上按照海試結果所繪製的船速油耗表相比，或是採用單一類神經網路所預測的油耗相比，採用複合式類神經網路可依照五種不同模型的預測結果設定適當配比，避免由於單一模型預測失準造成偏差。;By 2023, vessels in international trade must comply with IMO (International Maritime Organization) EEXI (Energy Efficiency Index for Existing Ships) and CII (Carbon Intensity Index) requirements which are aiming at a significant reduction of both total carbon emission and carbon emission intensity. IMO has further set up a strategy that will request vessels to reduce carbon intensity by 40% in 2030 and total emission by 70% in 2050 compared to the 2008 baseline. Therefore, vessel operators are now seeking possible solutions that could achieve this emission goal. Before 2030, most vessels are improving optimized hull design and speed adjustment (operation). Speed adjustment should be based on the precise prediction of main engine fuel oil consumption under different speeds and engine working conditions. By applying the 5-model hybrid artificial neural network and an adequate proportion of each model in the mixture, one can reduce residuals in fuel oil consumption (FOC) prediction in comparison with the FOC-speed table obtained from sea trial or normal single model prediction.
显示于类别:	[機械工程學系碩士在職專班 ] 博碩士論文

文件中的档案:

档案	描述	大小	格式	浏览次数
index.html		0Kb	HTML	76	检视/开启

在NCUIR中所有的数据项都受到原著作权保护.

社群 sharing

数据加载中.....