Offshore Wind Data Integration

Abstract

Using renewable energy to meet the future electricity consumption and to reduce environmental impact is a significant target of many countries around the world. Wind power is one of the most promising renewable energy technologies. In particular, the development of offshore wind power is increasing rapidly due to large areas of wind resources. However, offshore wind is encountering big challenges such as effective use of wind power plants, reduced cost of installation as well as operation and maintenance (O&M). Improved O&M is likely to reduce the hazard exposure of the employees, increase income, and support offshore activities more efficiently. In order to optimize the O&M, the importance of data exchange and knowledge sharing within the offshore wind industry must be realized. With more data available and accessible, it is possible to make better decisions, and thereby improve the recovery rates and reduce the operational costs. This dissertation proposes a holistic way of improving remote operations for offshore wind farms by using data integration. Particularly, semantics and integration aspects of data integration are investigated. The research looks at both theoretical foundations and practical implementations. As the outcome of the research, a framework for data integration of offshore wind farms has been developed. The framework consists of three main components: the semantic model, the data source handling, and the information provisioning. In particular, an offshore wind ontology has been proposed to explore the semantics of wind data and enable knowledge sharing and data exchange. The ontology is aligned with semantic sensor network ontology to support management of metadata in smart grids. That is to say, the ontology-based approach has been proven to be useful in managing data and metadata in the offshore wind and in smart grids. A quality-based approach is proposed to manage, select, and provide the most suitable data source for users based upon their quality requirements and an approach to formally describing derived data in ontologies is investigated.