Development a Typhoon-Induced Wave Forecasting Model and Extreme Value Analysis for Offshore Wind Farm Using Machine Learning Theory(II)

View Statistics Email Alert RSS Feed

Information

Details

Project title

Code/計畫編號

MOST108-2218-E019-001-MY2

Translated Name/計畫中文名

利用機器學習理論發展離岸風場之颱風波浪預測模式與極值分析(II)

Project Coordinator/計畫主持人

Wei-Ting Chao

Funding Organization/主管機關

National Science and Technology Council

Department/Unit

Center of Excellence for Ocean Engineering

Website

https://www.grb.gov.tw/search/planDetail?id=13327857

Year

2020

Start date/計畫起

01-08-2020

Expected Completion/計畫迄

31-07-2021

Bugetid/研究經費

635千元

ResearchField/研究領域

土木水利工程

Accurate and efficient prediction of typhoon-induced wave is an important research topic in oceanic science and offshore wind energy application. The main purpose of this project is to develop a wave forecasting model and extreme value analysis in typhoon conditions for offshore wind farm using machine learning theory. This study aims to gain deeper understanding of the physical characteristics of wind waves under typhoon conditions in coastal waters around Changhua county, which apply to development, operation and maintenance for the offshore wind farm. First, a detailed analysis will be conducted to characterize the statistical of marine meteorological data. Second, effective controlling parameters will be carefully investigated, including the center air pressure, the maximum wind speed, the radius of the typhoon, the distance between the station and typhoon center, as well as the translation speed and direction of the typhoon. In addition, the bathymetry (i.e., water depth and bottom slope) at the gauge station will be also considered to reflect the nearshore effect of the typhoon waves. Third, typhoon cyclone physics will be used in conjunction with several machine learning theories. The combined effective information can not only reduce the input dimension but also improve the model's learning and forecasting capability. Last, a suitable parametric typhoon model in coastal waters around Taiwan is developed by including the water depth and bottom slope of gauge station. Overall, this study attempts to push the artificial intelligence prediction of typhoon waves to a new milestone. In the future, a series of further research will be systematically carried out. The prediction results of typhoon-induced waves are expected to achieve significant efficiency and accuracy. 颱風波浪主要由颱風低壓及複雜的風場所產生，近年來在科學以及離岸風能應用上成為重要的研究議題。本計畫主要目的是利用機器學習理論發展離岸風場之颱風波浪預測模式與極值分析。計畫主要目的為透過深入的了解以及掌握彰濱離岸風場其周遭海域於颱風事件下風浪的物理特性，以作為離岸風場開發以及運維之用。計畫中將妥善分析觀測資料，以瞭解海氣象資料之統計特性；詳盡探討可能影響颱風波浪的因子(如颱風中心氣壓、最大風速、暴風半徑、距測站距離、颱風前進角度及方向)。此外，為了能建立一通用型之預報系統，本研究將測站之地形水深條件作為輸入因子，藉此反應颱風及波浪於近岸時，受到地形水深之影響；結合物理基礎與學習理論，整併結合有效的輸入資訊，減少輸入的維度、提昇學習與預測的效果；透過引入水深、地形參數，建立適合台灣鄰近海域之參數化風場模型。整體而言，本研究預期將人工智慧風浪預測模式推向新的里程碑。未來，可進一步進行一系列系統性研究課題，更詳盡瞭解掌握颱風事件下之風浪特性，並達到高效率且準確的颱風風浪預測結果。

Keyword(s)

Typhoon-induced waves
Effective controlling parameters
Parametric typhoon model
Machine learning theory
Extreme value analysis
颱風波浪
有效影響因子
參數化颱風模型
機器學習理論
極端值分析

DSpace CRIS

Development a Typhoon-Induced Wave Forecasting Model and Extreme Value Analysis for Offshore Wind Farm Using Machine Learning Theory(II)

Details

Description