Search

Browse

• Communities
• Authors
• Year
• Titles
• Subjects

My Space

• Login

Optimal siting of wave energy converters using machine learning and GIS

Batsis Georgios

Full record

---

Summary

Nowadays, Renewable Energy Sources are the solution to the problem of ever-increasing climate change. For this reason, several countries are proceeding with the electricity production using alternative sources such as solar and wind farms. This makes it possible to limit the operation of conventional means of production (lignite, coal-fired) gradually. One of the alternative sources of electricity production are the ocean waves. Wave energy Converters are systems that convert wave energy into electrical energy. It is alleged that the percentage of energy produced by renewable sources using wave energy can be increased on a large-scale in the future in regions such as Scandinavia, the Mediterranean, the United Kingdom, Oceania and the maritime region of the American continent.In this work, the optimal siting of the installation of wave energy converters is examined, which is one of the main areas of research of this field. It is essential to take into account geospatial and technical limitations, in order to find the optimal locations. Geospatial constraints depend on both seagrass and the Land Use classes of the closest coastal area. Technical limitations include meteorological conditions and the morphology of the seabed. Suitable installation areas are selected after the exclusion of points that do not meet the aforementioned restrictions.Searching for optimal locations is achieved through the utilization of Machine Learning algorithms. Initially, a Deep Neural Network that is implemented operates based on heterogeneous data fusion, in this case satellite images and time series of meteorological data. This fact implies the definition of two branches architecture. The branch that is trained with image data provides for the localization of dynamic geospatial classes in the potential installation area, whereas the second one is responsible classifier the region according to the potential wave energy using wave height and period time series. In making the final decision on the suitability of the candidate area, a large number of static land use data play an important role, the utilization of which does not require a Machine Learning algorithm. Therefore, these data are combined with neural network predictions for the optimizing positioning of the Wave Energy Converters.For the sake of completeness and flexibility, a Multi-Task Neural Network is developed. This model, in addition to predicting the suitability of an area depending on marine flora and wave energy, also predicts land use classes through Multi-Label classification process. In this case, the potential regions classified as suitable or as not suitable for the installation of Wave Energy Converter system exclusively with the help of neural network predictions.Obviously, a large amount of heterogeneous data is required for the development and implementation of the aforementioned system and Deep Neural Network. For this reason, a geographic information tool that aims to receive and store georeferenced data is developed. This tool is employed to develop training and validation datasets and to use the overall system in the desired case study.The proposed methodology is applied in the marine area of the city of Sines, Portugal. In this geographical area, they also focus on papers in which optimal zoning is carried out through the traditional methods. The present work includes, among others, potential nearshore areas. For this reason, the comparison of the results can be carried out mainly for the offshore points. Although the evaluation of wave energy is carried out through different methods and in the present work is calculated on the study and seagrass, there is agreement of the results largely. The first neural network achieves 98.7% Binary Classification accuracy, while the Multi-Task Neural Network 97.5 % in the same metric and 93.5% in the F1 score of the Multi-Label classification output.

Available Files

Services