Application of an active learning method for cumulative fatigue damage assessment of floating wind turbine mooring lines

Abstract

Long-term cumulative fatigue damage of mooring lines is crucial for the design of floating wind turbine structures (FWTs). Although many efforts are carried out for the offshore floating platforms, there still needs to be an efficient approach for assessing the long-term fatigue damage of the floating wind turbine mooring lines due to the complex loading in FWTs. An active learning approach named AK-DA (Adaptive Kriging Damage Assessment) was recently proposed for the cumulative fatigue damage assessment of wind turbine structures. However, in the original work, the AK-DA approach was only tested on fatigue damage assessment of a 5MW wind turbine tower with monopile support structures. It is unclear whether it is applicable to other parts of the wind turbine system, especially considering the complex loading of FWTs. Therefore, in this work, the AK-DA approach is used to assess the cumulative fatigue damage of mooring lines. The Gaussian process regression (Kriging) model is used to predict the fatigue damage of the mooring line under different wind-wave cases. The cumulative fatigue damage of the mooring lines in the IEA 15MW semi-submersible wind turbines is assessed with the AK-DA approach. The numerical simulation results show that the AK-DA approach can efficiently and accurately estimate the cumulative fatigue damage of the mooring lines. Compared to the traditional simulation approach, the AK-DA approach can increase efficiency by more than 45 times, and the absolute error is less than 1%. This active learning approach could serve as a helpful tool for offshore mooring system designers, facilitating the cumulative fatigue damage assessment during the design process.