Wind turbines are a crucial source of renewable energy, but their blades are constantly exposed to harsh weather conditions, which can cause damage over time. To ensure the safety and efficiency of wind turbines, it is essential to detect damage in the composite materials used to make the blades. However, detecting damage in composites can be a challenging task. Traditional methods of inspection are time-consuming, costly, and often not very accurate.
To address this issue, researchers have developed a new method for detecting damage in composites using Machine Learning. The method is based on predicting maximum strain values in an undamaged composite material and comparing these values to the measured values in a damaged composite material in a region under investigation. A change or anomaly in the strains collected from these regions that may be caused by the presence of damage will be identified by the Machine Learning model.
The algorithm was trained using the displacement values (U) at every time step and the distance of each region (Δx, Δy) to the applied loading. The researchers introduced an elliptical defect in some of the regions in the composite materials and analyzed the effect of the damage on the strain states. The presence of damage was found to have a high influence on the strain states of the material.
Supervised learning models such as Linear Regression (LR), Random Forest (RF) and Artificial Neural Networks (ANN) were used to predict the strain values in different regions based on the displacement and distance of the region to the applied loading. The LR model achieved the highest accuracy at 89%, followed by RF at 79% and ANN at 56%.
This research shows that Machine Learning can be used effectively in the prediction of damages or defects in composite materials used in wind turbine blades. The models achieved good accuracy, but there is still room for improvement. The researchers suggested that an approach to improve the model will be to generate more data by introducing several loading conditions. This will allow for more accurate and efficient detection of damages in composite materials, ensuring the safety and reliability of wind turbines.
