Publication
Artificial Intelligence for the Prediction of Faillures in Wind Turbines
| Summary: | With the increase of the global environmental awareness and demand nowadays, the request for renewable energy sources by the companies responsible for distributing the energy itself highly increased as well. The sources, such as wind turbines, are highly exposed to several external factors that can result in mechanical failures on their complex components. These faults lead to multiple consequences due to their failure time: loss of energy production, which means that less renewable energy will be transmitted to the electrical grid and more money will be spent on repairing these components. The content of this thesis consists of an analysis to the implementation of a prediction system allied with a monitoring system. This system will allow failure detection on wind assets, anticipating them and thus reducing maintenance costs and the increase of the longevity of the components. By reducing the number of failures, it will allow companies to increase the profit of the energy production on a long term. With the work that was made, it was implemented an experiment that could be applied to several faults of wind turbines that created machine learning models that reliably predict these faults. |
|---|---|
| Main Authors: | Rebimba, Diogo Jorge Gato |
| Subject: | Renewable energy machine learning renewable monitoring system wind parks wind turbine failure prediction |
| Year: | 2022 |
| Country: | Portugal |
| Document type: | master thesis |
| Access type: | open access |
| Associated institution: | Universidade Nova de Lisboa |
| Language: | English |
| Origin: | Repositório Institucional da UNL |
| Summary: | With the increase of the global environmental awareness and demand nowadays, the request for renewable energy sources by the companies responsible for distributing the energy itself highly increased as well. The sources, such as wind turbines, are highly exposed to several external factors that can result in mechanical failures on their complex components. These faults lead to multiple consequences due to their failure time: loss of energy production, which means that less renewable energy will be transmitted to the electrical grid and more money will be spent on repairing these components. The content of this thesis consists of an analysis to the implementation of a prediction system allied with a monitoring system. This system will allow failure detection on wind assets, anticipating them and thus reducing maintenance costs and the increase of the longevity of the components. By reducing the number of failures, it will allow companies to increase the profit of the energy production on a long term. With the work that was made, it was implemented an experiment that could be applied to several faults of wind turbines that created machine learning models that reliably predict these faults. |
|---|