A Multivariate Spatiotemporal Feature Fusion Network for Wind Turbine Gearbox Condition Monitoring

SCADA data, due to their easy accessibility and low cost, have been widely applied in wind turbine gearbox condition monitoring. However, the high-dimensional and nonlinear nature of the collected data, along with the insufficient spatiotemporal feature capabilities of existing methods and the lack...

Full description

Saved in:
Bibliographic Details
Main Authors: Shixian Dai, Shuang Han, Xinjian Bai, Zijian Kang, Yongqian Liu
Format: Article
Language:English
Published: MDPI AG 2025-03-01
Series:Energies
Subjects:
Online Access:https://www.mdpi.com/1996-1073/18/5/1273
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850050812472983552
author Shixian Dai
Shuang Han
Xinjian Bai
Zijian Kang
Yongqian Liu
author_facet Shixian Dai
Shuang Han
Xinjian Bai
Zijian Kang
Yongqian Liu
author_sort Shixian Dai
collection DOAJ
description SCADA data, due to their easy accessibility and low cost, have been widely applied in wind turbine gearbox condition monitoring. However, the high-dimensional and nonlinear nature of the collected data, along with the insufficient spatiotemporal feature capabilities of existing methods and the lack of consideration of the physical mechanisms of wind turbine operation, limit the accuracy of monitoring models. In this paper, a multivariate spatiotemporal feature fusion network is proposed for wind turbine gearbox condition monitoring. First, by analyzing the operational mechanism of wind turbines and the correlation between sensor data, the time series data are transformed into graph data. Then, graph convolutional networks and temporal convolutional networks are used to extract spatial and temporal features, respectively. Next, long short-term memory networks are employed to fuse the extracted temporal and spatial features, further capturing long-term spatiotemporal dependencies. Finally, the proposed method is validated using real data from two wind turbines. Experimental results show that the proposed method reduces the RMSE by 29.67% and 17.61% compared to the best-performing models. Moreover, the proposed method provides early warning signals 188.6 h and 133.67 h in advance, achieving stable and efficient early anomaly detection for wind turbines.
format Article
id doaj-art-861cca4eeecc4e62af6726d513d036c0
institution DOAJ
issn 1996-1073
language English
publishDate 2025-03-01
publisher MDPI AG
record_format Article
series Energies
spelling doaj-art-861cca4eeecc4e62af6726d513d036c02025-08-20T02:53:19ZengMDPI AGEnergies1996-10732025-03-01185127310.3390/en18051273A Multivariate Spatiotemporal Feature Fusion Network for Wind Turbine Gearbox Condition MonitoringShixian Dai0Shuang Han1Xinjian Bai2Zijian Kang3Yongqian Liu4State Key Laboratory of Alternate Electrical Power System with Renewable Energy Sources, School of New Energy, North China Electric Power University, Beijing 102206, ChinaState Key Laboratory of Alternate Electrical Power System with Renewable Energy Sources, School of New Energy, North China Electric Power University, Beijing 102206, ChinaState Key Laboratory of Alternate Electrical Power System with Renewable Energy Sources, School of New Energy, North China Electric Power University, Beijing 102206, ChinaState Key Laboratory of Alternate Electrical Power System with Renewable Energy Sources, School of New Energy, North China Electric Power University, Beijing 102206, ChinaState Key Laboratory of Alternate Electrical Power System with Renewable Energy Sources, School of New Energy, North China Electric Power University, Beijing 102206, ChinaSCADA data, due to their easy accessibility and low cost, have been widely applied in wind turbine gearbox condition monitoring. However, the high-dimensional and nonlinear nature of the collected data, along with the insufficient spatiotemporal feature capabilities of existing methods and the lack of consideration of the physical mechanisms of wind turbine operation, limit the accuracy of monitoring models. In this paper, a multivariate spatiotemporal feature fusion network is proposed for wind turbine gearbox condition monitoring. First, by analyzing the operational mechanism of wind turbines and the correlation between sensor data, the time series data are transformed into graph data. Then, graph convolutional networks and temporal convolutional networks are used to extract spatial and temporal features, respectively. Next, long short-term memory networks are employed to fuse the extracted temporal and spatial features, further capturing long-term spatiotemporal dependencies. Finally, the proposed method is validated using real data from two wind turbines. Experimental results show that the proposed method reduces the RMSE by 29.67% and 17.61% compared to the best-performing models. Moreover, the proposed method provides early warning signals 188.6 h and 133.67 h in advance, achieving stable and efficient early anomaly detection for wind turbines.https://www.mdpi.com/1996-1073/18/5/1273wind turbine gearboxgraph convolutional networktemporal convolutional networklong short-term memory networkanomaly detection
spellingShingle Shixian Dai
Shuang Han
Xinjian Bai
Zijian Kang
Yongqian Liu
A Multivariate Spatiotemporal Feature Fusion Network for Wind Turbine Gearbox Condition Monitoring
Energies
wind turbine gearbox
graph convolutional network
temporal convolutional network
long short-term memory network
anomaly detection
title A Multivariate Spatiotemporal Feature Fusion Network for Wind Turbine Gearbox Condition Monitoring
title_full A Multivariate Spatiotemporal Feature Fusion Network for Wind Turbine Gearbox Condition Monitoring
title_fullStr A Multivariate Spatiotemporal Feature Fusion Network for Wind Turbine Gearbox Condition Monitoring
title_full_unstemmed A Multivariate Spatiotemporal Feature Fusion Network for Wind Turbine Gearbox Condition Monitoring
title_short A Multivariate Spatiotemporal Feature Fusion Network for Wind Turbine Gearbox Condition Monitoring
title_sort multivariate spatiotemporal feature fusion network for wind turbine gearbox condition monitoring
topic wind turbine gearbox
graph convolutional network
temporal convolutional network
long short-term memory network
anomaly detection
url https://www.mdpi.com/1996-1073/18/5/1273
work_keys_str_mv AT shixiandai amultivariatespatiotemporalfeaturefusionnetworkforwindturbinegearboxconditionmonitoring
AT shuanghan amultivariatespatiotemporalfeaturefusionnetworkforwindturbinegearboxconditionmonitoring
AT xinjianbai amultivariatespatiotemporalfeaturefusionnetworkforwindturbinegearboxconditionmonitoring
AT zijiankang amultivariatespatiotemporalfeaturefusionnetworkforwindturbinegearboxconditionmonitoring
AT yongqianliu amultivariatespatiotemporalfeaturefusionnetworkforwindturbinegearboxconditionmonitoring
AT shixiandai multivariatespatiotemporalfeaturefusionnetworkforwindturbinegearboxconditionmonitoring
AT shuanghan multivariatespatiotemporalfeaturefusionnetworkforwindturbinegearboxconditionmonitoring
AT xinjianbai multivariatespatiotemporalfeaturefusionnetworkforwindturbinegearboxconditionmonitoring
AT zijiankang multivariatespatiotemporalfeaturefusionnetworkforwindturbinegearboxconditionmonitoring
AT yongqianliu multivariatespatiotemporalfeaturefusionnetworkforwindturbinegearboxconditionmonitoring