Rough Set Theory Based Blade Condition Classification on Wind Turbine through Statistical Features

Abstract

Wind energy has become a one of the alternative energy source due to fossil fuel crisis. These wind energies are being harvested from the wind through wind turbines. These wind turbines are subjected to various environmental factors and prone to severe vibration on blade. This vibration lead to the catastrophic calamities and cause severe capital loss and wind production loss. This study proposes a data processing and analysis of wind turbine blade faults using rough set theory based feature classification. The feature extraction (statistical features) and the feature selection (J48 decision tree algorithm) methods were used to identify the best features for fault classification. Using rough set theory, with five statistical features, 75.5% of classification accuracy have been obtained for the fault identification on wind turbine blade.