A surrogate model for estimating extreme tower loads on wind turbines based on random forest proximities

Area of Science:

• Engineering
• Renewable Energy
• Computational Science

Background:

• Assessing extreme tower loads is crucial for wind turbine design, as mandated by International Electrotechnical Commission (IEC) Standard 61400-1.
• Traditional methods may present limitations in accurately capturing probabilistic characteristics of loads.
• There is a need for robust surrogate models to efficiently estimate these critical loads.

Purpose of the Study:

• To present a novel surrogate model for estimating extreme tower loads on wind turbines.
• To provide a method that complies with IEC Standard 61400-1 requirements for wind turbine design.
• To develop an adaptable model for high-dimensional and sparse data settings.

Main Methods:

• A surrogate model is developed using random forests to impute tower loads.
• The model matches observed operational signals with simulated quantities.
• Proximities induced by random forests are utilized for load estimation, avoiding regression-based approaches.

Main Results:

• The proposed surrogate model effectively estimates extreme tower loads.
• The model demonstrates adaptability to high-dimensional and sparse data.
• Application to an operating wind turbine validates the model's performance using operational statistics.

Conclusions:

• The developed surrogate model offers a reliable method for estimating extreme tower loads in wind turbines.
• This approach enhances the design phase by providing accurate load estimations compliant with industry standards.
• The model's ability to handle complex data makes it a valuable tool for wind energy research and development.