Detecting Anomaly Classification Using PCA-Kmeans and Ensembled Classifier for Wind Turbines

Abstract

Monitoring wind turbine performance is vital for ensuring wind turbines’ safe, efficient, and cost-effective operation over time. Using principal component analysis (PCA), k-means clustering for labeling, and an ensemble classifier for finding outliers, this study suggests a new way to find anomalies in wind turbines. The primary objective is to improve the precision of anomaly detection in wind turbines by leveraging machine-learning techniques. The proposed methodology utilizes the output of the PCA-Kmeans model to label supervisory control and data acquisition (SCADA) data. Furthermore, a stacking ensemble classifier is employed to refine the model’s precision. Our proposed model achieved a classification accuracy of 99%, which is a significant improvement compared to existing approaches. The significance of this study lies in its potential to enable more efficient wind turbine operation by identifying and resolving anomalies that may reduce their performance. This can ultimately contribute to achieving a sustainable and renewable energy future.

Keywords

• Wind turbine
• anomaly detection
• principal component analysis
• k-means clustering
• labeling
• ensemble classifier