Predicting Remaining Useful Life of Wind Turbine Bearing using Linear Regression

Abstract

Almost all industrial wind turbine failures are caused by bearing degeneration. As a critical part of this functionality, precisely estimating the remaining usable life (RUL) of the bearings is necessary in order to ensure the reliability and availability of energy generation. This work investigates how to build a classification model in Python to estimate the RUL of a wind turbine main bearing. Making use of SCADA data Given by the Harvard Dataverse data set, we select only the four most physical characteristics from this data set to solve this challenge. Temperature, viscosity, dynamic load, and fatigue damage are all factors to consider. The suggested methods is based on a concept that was previously developed in the literature for prediction in other discipline. This paper also assesses which model is best for forecasting failure on the given data set. This assessment is carried out in order to determine that linear regression is the best method for producing a model capable of reducing variation and improving the metrics of our model with an accuracy level of 99 percent for daily prediction. This enables the development of a new sort of intelligent intention.