Abstract:The predictive maintenance of wind turbines has become a critical issue with the rapid development of wind power generation. The early detection of abnormal operation conditions can prevent failure status, which takes a long time to recover. Energy waste can also be reduced while maintenance efficiency can be improved by using a supervisory control and data acquisition (SCADA) system to monitor the operation status of wind turbines. Massive data are generated from different sensors during wind turbine operation, and SCADA can be used to gather reports about hundreds of possible abnormal conditions. The popular maintenance methods have been mostly designed on the basis of statistical analysis and data mining. However, such schemes need not only big data but also sophisticated processing techniques. This study addresses the aforementioned challenges by proposing a deep learning model with comprehensive data preprocessing and hyperparameter tuning on batch size to achieve abnormal early detection. The necessary data preprocessing is initially conducted besides the conventional data cleaning and normalization steps, and time-series data windowing and label settings are also performed. Then, the imbalanced classes in the records are addressed by adopting an augmentation scheme called the synthetic minority oversampling technique. Principal component analysis is also used to enhance the training. Finally, the proposed deep learning method with fine-tuning is compared with three machine learning models for early anomaly event detection. Experimental results show that the proposed scheme can identify potential faults 72 hours before they occur, and the precision rate exceeds 90 %.

A Semi-Supervised Anomaly Detection Method For Wind Farm Power Data Preprocessing

An Adaptive Identification Method of Abnormal Data in Wind and Solar Power Stations

Anomaly Detection Algorithm Based on Electric Equipment

An Improved Data-Efficiency Algorithm Based on Combining Isolation Forest and Mean Shift for Anomaly Data Filtering in Wind Power Curve

Enhanced Anomaly Detection and Normal Behaviour Power Curve Modelling in Wind Farm SCADA Data: A Hybrid Approach

Semi-supervised multivariate time series anomaly detection for wind turbines using generator SCADA data

A two-stage ensemble of diverse models for recognition of abnormal data in raw wind data

Multivariate Anomaly Detection and Early Warning Framework for Wind Turbine Condition Monitoring Using SCADA Data

Abnormal Data Processing of Wind Turbine Based on Combined Algorithm and Class Center Imputation

Image-Based Abnormal Data Detection and Cleaning Algorithm via Wind Power Curve

Anomaly detection for wind turbines based on the reconstruction of condition parameters using stacked denoising autoencoders

Study of Wind Turbine Fault Diagnosis and Early Warning Based on SCADA Data

Anomaly detection and diagnosis for wind turbines using long short-term memory-based stacked denoising autoencoders and XGBoost

Wind Power Curve Data Cleaning by Image Thresholding Based on Class Uncertainty and Shape Dissimilarity

An unsupervised spatiotemporal graphical modeling approach for wind turbine condition monitoring

Semi-Supervised Domain Adaptation with Feature Separation for Wind Turbine Anomaly Detection

Abnormal Data Cleaning for Wind Turbines by Image Segmentation Based on Active Shape Model and Class Uncertainty

Data-Driven Day-ahead Probabilistic Forecasting of Wind Power Based on Features Sensitivity Analysis and Meteorological Scenario Classification

Anomaly Detection of Wind Turbines Based on Deep Small-World Neural Network

Method for Cleaning Abnormal Data of Wind Turbine Power Curve Based on Density Clustering and Boundary Extraction

Predictive maintenance of abnormal wind turbine events by using machine learning based on condition monitoring for anomaly detection