Tackling the Reliability Problem on Loads Prediction in Wind Turbines
Reducing risk and efficiently manage lifetime of the wind turbines is key. The wind industry has found an ally in machine learning to get to know and efficiently manage the life of a wind farm. Combining those tools with real measurements we can tackle the problem of the reliability when predicting the loads for the different components of the wind turbine.
Wind turbine manufacturers, windfarm owners and operators need better knowledge about real loads in the main components of their wind turbines in order to run their wind farms safely and in the most profitable way at the same time. Therefore, DNV GL is currently developing LoadsPredict, a tool that can significantly contribute to the optimization between power performance and loads with an affordable and qualified solution based on real measurement data.
Our skilled engineers use load measurements to create neural network models capable of determining complex, nonlinear relations and patterns across and between measured quantities allowing them to add value by giving precise information for better performance, manage life-time, data of an unknown past and predicting the future.
Machine learning techniques combined with measurement campaigns enables to improve model-based tools optimising performance results and reducing risk of failure.
