Improvement in wind energy production through condition monitoring of wind turbine blades using vibration signatures and ARMA features: a data-driven approach

The main objective of this study is to improve the wind energy productivity by implementing the condition monitoring technique for wind turbine blades through vibration source. The fault detection and the isolation of the fault which affects the wind energy productivity were carried using machine learning algorithms. In this study, a three bladed horizontal axis wind turbine was chosen and the faults like blade bend, blade cracks, hub-blade loose connection, blade erosion and pitch angle twist were considered as these are the faults which affect the turbine blade. Initially, vibration sources were collected from the wind turbine using piezoelectric accelerometer and from that vibration source; needed features are extracted using ARMA through MATLAB. From the extracted feature, the dominating feature is selected using J48 decision tree algorithm and with the selected features, fault classification has been carried out. The fault classifications were carried out using Bayesian, function and lazy classifiers.