Abstract Mediterranean islands have the advantage of favourable climatic conditions to use different marine renewable energy sources. Remote sensing can provide data to determine wind energy production potential and observational activity to identify, assess and detect suitable points in large marine areas. In this paper, a new combined model has been developed to integrate wind speed assessment, mapping and forecasting using Sentinel 1 satellite data through images processing and Adaptive Neuro-Fuzzy Inference System and the Bat algorithm. Synthetic Aperture Radar (SAR) satellite images from the Sentinel 1 satellite have been used in order to detect offshore and nearshore wind potential. Particularly, Sentinel 1 images have been analysed by means of the SNAP software. Then, to extract data about wind speed and direction, a GIS software for mapping the wind climate has been used. This new methodology has been applied to the North-Central coasts of Sardinia Island and then focused on six main small islands of La Maddalena archipelago. Furthermore, ten Hot Spots (HSs) have been identified as interesting because of their high-energy potential and the possibility to be considered as sites for future implementation of Wind Turbine Generators (WTGs). Finally, the ten identified HS have been used as input data to train and test the proposed forecast model.

中文翻译：

---

使用 Sentinel 1 卫星和基于机器学习的新预测模型进行风源潜力评估：以撒丁岛为例

摘要 地中海岛屿具有有利的气候条件，可以利用不同的海洋可再生能源。遥感可以提供数据来确定风能生产潜力和观测活动，以识别、评估和检测大型海洋区域的合适点。在本文中，开发了一种新的组合模型，通过图像处理和自适应神经模糊推理系统和 Bat 算法，使用 Sentinel 1 卫星数据集成风速评估、测绘和预测。来自哨兵 1 号卫星的合成孔径雷达 (SAR) 卫星图像已被用于检测海上和近岸风电潜力。特别是，Sentinel 1 图像已通过 SNAP 软件进行分析。然后，为了提取有关风速和风向的数据，已使用用于绘制风气候图的 GIS 软件。这种新方法已应用于撒丁岛的中北部海岸，然后专注于拉马达莱纳群岛的六个主要小岛。此外，十个热点 (HS) 已被确定为有趣的，因为它们具有高能量潜力，并且有可能被视为未来实施风力涡轮发电机 (WTG) 的地点。最后，十个识别出的 HS 已被用作输入数据来训练和测试所提出的预测模型。十个热点 (HS) 已被确定为有趣的，因为它们具有高能量潜力，并且有可能被视为未来实施风力涡轮发电机 (WTG) 的地点。最后，十个识别出的 HS 已被用作输入数据来训练和测试所提出的预测模型。十个热点 (HS) 已被确定为有趣的，因为它们具有高能量潜力，并且有可能被视为未来实施风力涡轮发电机 (WTG) 的地点。最后，十个识别出的 HS 已被用作输入数据来训练和测试所提出的预测模型。