Wind Turbine Bearing Anomaly Detection Using CMMS Data and Machine Learning

EasyChair Preprint 15256

Abstract

The need to anticipate failures in wind turbines has become increasingly urgent. The exponential increase in the number of installed turbines, coupled with the aging of the generation fleet, has intensified the competition to reduce operation and maintenance costs, which means minimizing unplanned downtime and costly major repairs. The aim of this study is to utilize the vibration data available in the CMMS to identify turbines with significant condition deviations that pose a high risk of failure. The data processing approach using CNN and PCA in the pre-processing stage, along with SVM for health state classification, demonstrated excellent accuracy for both single turbine and multi-turbine tests, making it suitable for managing wind farms with a large number of turbines.

Keyphrases: Bearing, Condensed Nearest Neighbor (CNN), Condition Monitoring Management System (CMMS), Principal Component Analysis (PCA), Support Vector Machine (SVM), machine learning, wind turbine

@booklet{EasyChair:15256,
  author    = {Gabriel Freitas Santos and Helon Vicente Hultmann Ayala and Florian Alain Yannick Pradelle and Paula Aida Sesini},
  title     = {Wind Turbine Bearing Anomaly Detection Using CMMS Data and Machine Learning},
  howpublished = {EasyChair Preprint 15256},
  year      = {EasyChair, 2024}}