Failure Mode Classification of Wind Turbine Gearbox Utilizing Machine Learning Algorithms

There has been a substantial expansion in the use of renewable resources, such as wind energy. Wind energy is converted into electricity, which is situated outside, rendering them vulnerable to inclement weather situations. These situations can result in various forms of harm to wind turbines, reducing their lifespan and efficiency and subsequently increasing their operational expenses. Hence, monitoring conditions and identifying problems early are of utmost importance. By employing machine learning techniques, it is possible to proactively avoid the deterioration of wind turbine components and identify and diagnose malfunctions. This method enables the automated and self-directed knowledge acquisition to forecast, identify, and analyze mechanical malfunctions in wind turbines. The study simulated three distinct failure scenarios in wind turbines and compared the performance of various AI methodologies, including Extra tree, Random Forest, Decision Tree, Neural Networks, and SVM. This work assesses various approaches for statistical feature extraction, such as standard deviation, mean, kurtosis, and skewness. Implementing these strategies makes it possible to predict and classify equipment failure types. Significant outcomes are achieved with a precision of 96.6%.