Posters

Abstract

Knowing where offshore wind infrastructures are located offers the data basis to various insights on offshore wind farm installation when combined with industry knowledge and experience to ensure the correct interpretation. Then such data provides valuable knowledge on wind turbine generator (WTG) installation times, creates the basis for mapping portals or decision support systems. However, a significant challenge in gaining such insights may be the lack of data on the position of individual WTG. In many regions, such data may not be publicly available due to amongst other the lack of an appropriate spatial data infrastructure, lack of public interest, limited interest of project developers sharing such information, or for political reasons. Hence, the aim of this study is to find an effective and accurate method of determining WTG positions [MZ1] based on available public data. For this purpose, the Algorithm for Offshore Wind Infrastructure Detection (AOWID) algorithm is being proposed. The algorithm composes of three variants, all based on unsupervised machine learning techniques, that have been tested for three selected wind farms for which reference data has been obtained: Eneco Luchterduinen, Race Bank and Horns Rev 3.