Mapeamento de métodos com Digital Twin em turbinas eólicas offshore

Abstract

In recent decades, countries have intensified efforts in the development and adopt strategies for the use of clean energy, aiming to reduce environmental impacts and promote planetary sustainability. Offshore wind energy stands out as it presents significant potential for electricity generation due to its efficiency and the consistency of wind resources at sea. However, the logistics and maintenance of these structures involve high costs, posing a significant challenge to the economic viability of the projects. As a result, optimizing the operation and maintenance of offshore wind farms has become a growing area of interest for the industrial sector, which seeks ways to enhance efficiency and reduce operational costs. Real-time analysis and prediction methods for the components of offshore wind turbines are constantly being developed and improved, incorporating new technologies and approaches. The digital twin aims to monitor and support decision-making by project stakeholders, providing accurate and up-to-date data, while ensuring predictive accuracy, thereby contributing to efficient asset management. In this context, this paper proposes to map the methods used to obtain a digital twin, with a focus on offshore wind turbines. The research methodology involved a systematic literature review and a content analysis about digital twin methods.. The results revealed that the Finite Element Method (FEM) is the most widely used analysis method (31%), while Reduced-Order Models (ROM) are the most commonly applied prediction methods (20%). Furthermore, it was found that the choice and application of analysis or prediction methods may vary depending on the specific application and component. Digital Twin has diverse applications and methods for achieving final results, wherein signals generated by the physical object are mathematically modeled for use in prediction methods and fed back into the cycle to assist in decision-making.