Data-Driven Intelligent Prediction Model for Gearbox Temperature in Wind Turbines

Wind power has become increasingly popular due to the demand for sustainable energy sources. A fundamental problem these renewable sources face is the energy cost, thus challenging their efficient operation and maintenance (O&M). Predictive maintenance (PM) has been developed to be an economical solution for the active overcoming of problems, such as wind turbine (WT) gearbox bearing failures that seriously affect the performance and stability of renewable energy systems (RES). This study develops a predictive model of gearbox-bearing temperature using advanced machine learning algorithms such as multiple linear regression (MLR), Extreme Gradient boosting (XGBoost), and Random forest (RF) from the WT supervisory control and data acquisition (SCADA) system. The RF model performed better than the MLR and XGBoost model, with an R-squared value of 0.9840, and the Root mean square error (RMSE) and the Mean absolute percentage error (MAPE) were discovered to be 0.7352 and 0.88%, respectively.