Wind turbine condition monitoring based on three fitted performance curves
Abstract Based on SCADA data, this study aims at fitting three performance curves (PCs), power curve, pitch angle curve, and rotor speed curve, to accurately describe the normal behaviour of a wind turbine (WT) for performance monitoring and identification of anomalous signals. The fitting accuracy...
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Wiley
2024-05-01
|
| Series: | Wind Energy |
| Subjects: | |
| Online Access: | https://doi.org/10.1002/we.2859 |
| _version_ | 1797197295316369408 |
|---|---|
| author | Shuo Zhang Emma Robinson Malabika Basu |
| author_facet | Shuo Zhang Emma Robinson Malabika Basu |
| author_sort | Shuo Zhang |
| collection | DOAJ |
| description | Abstract Based on SCADA data, this study aims at fitting three performance curves (PCs), power curve, pitch angle curve, and rotor speed curve, to accurately describe the normal behaviour of a wind turbine (WT) for performance monitoring and identification of anomalous signals. The fitting accuracy can be undesirably affected by erroneous SCADA data. Hence, outliers generated from raw SCADA data should be removed to mitigate the prediction inaccuracy, so various outlier detection (OD) approaches are compared in terms of area under the curve (AUC) and mean average precision (mAP). Among them, a novel unsupervised SVM‐KNN model, integrated by support vector machine (SVM) and k nearest neighbour (KNN), is the optimum detector for PC refinements. Based on the refined data by the SVM‐KNN detector, several common nonparametric regressors have largely improved their prediction accuracies on pitch angle and rotor speed curves from roughly 86% and 90.6%, respectively, (raw data) to both 99% (refined data). Noticeably, under the SVM‐KNN refinement, the errors have been reduced by roughly five times and 10 times for pitch angle and rotor speed predictions, respectively. Ultimately, bootstrapped prediction interval is applied to conduct the uncertainty analysis of the optimal predictive regression model, reinforcing the performance monitoring and anomaly detection. |
| first_indexed | 2024-04-24T06:41:42Z |
| format | Article |
| id | doaj.art-ecc5f72889624504a14d6fae37a2b671 |
| institution | Directory Open Access Journal |
| issn | 1095-4244 1099-1824 |
| language | English |
| last_indexed | 2024-04-24T06:41:42Z |
| publishDate | 2024-05-01 |
| publisher | Wiley |
| record_format | Article |
| series | Wind Energy |
| spelling | doaj.art-ecc5f72889624504a14d6fae37a2b6712024-04-23T02:09:56ZengWileyWind Energy1095-42441099-18242024-05-0127542944610.1002/we.2859Wind turbine condition monitoring based on three fitted performance curvesShuo Zhang0Emma Robinson1Malabika Basu2Technological University Dublin Dublin IrelandTechnological University Dublin Dublin IrelandTechnological University Dublin Dublin IrelandAbstract Based on SCADA data, this study aims at fitting three performance curves (PCs), power curve, pitch angle curve, and rotor speed curve, to accurately describe the normal behaviour of a wind turbine (WT) for performance monitoring and identification of anomalous signals. The fitting accuracy can be undesirably affected by erroneous SCADA data. Hence, outliers generated from raw SCADA data should be removed to mitigate the prediction inaccuracy, so various outlier detection (OD) approaches are compared in terms of area under the curve (AUC) and mean average precision (mAP). Among them, a novel unsupervised SVM‐KNN model, integrated by support vector machine (SVM) and k nearest neighbour (KNN), is the optimum detector for PC refinements. Based on the refined data by the SVM‐KNN detector, several common nonparametric regressors have largely improved their prediction accuracies on pitch angle and rotor speed curves from roughly 86% and 90.6%, respectively, (raw data) to both 99% (refined data). Noticeably, under the SVM‐KNN refinement, the errors have been reduced by roughly five times and 10 times for pitch angle and rotor speed predictions, respectively. Ultimately, bootstrapped prediction interval is applied to conduct the uncertainty analysis of the optimal predictive regression model, reinforcing the performance monitoring and anomaly detection.https://doi.org/10.1002/we.2859area under the curve (AUC)bootstrapped prediction intervalmean average precision (mAP)outlier detection (OD)performance curve (PC)SVM‐KNN |
| spellingShingle | Shuo Zhang Emma Robinson Malabika Basu Wind turbine condition monitoring based on three fitted performance curves Wind Energy area under the curve (AUC) bootstrapped prediction interval mean average precision (mAP) outlier detection (OD) performance curve (PC) SVM‐KNN |
| title | Wind turbine condition monitoring based on three fitted performance curves |
| title_full | Wind turbine condition monitoring based on three fitted performance curves |
| title_fullStr | Wind turbine condition monitoring based on three fitted performance curves |
| title_full_unstemmed | Wind turbine condition monitoring based on three fitted performance curves |
| title_short | Wind turbine condition monitoring based on three fitted performance curves |
| title_sort | wind turbine condition monitoring based on three fitted performance curves |
| topic | area under the curve (AUC) bootstrapped prediction interval mean average precision (mAP) outlier detection (OD) performance curve (PC) SVM‐KNN |
| url | https://doi.org/10.1002/we.2859 |
| work_keys_str_mv | AT shuozhang windturbineconditionmonitoringbasedonthreefittedperformancecurves AT emmarobinson windturbineconditionmonitoringbasedonthreefittedperformancecurves AT malabikabasu windturbineconditionmonitoringbasedonthreefittedperformancecurves |