Condition monitoring and performance forecasting of wind turbines based on denoising autoencoder and novel convolutional neural networks

With the proportion of wind power in the grid increasing, the monitoring and maintenance of wind turbines is becoming more and more important for the reliability of the grid. In this study, a data-driven modelling framework based on deep convolutional neural networks is constructed for wind turbines...

Full description

Bibliographic Details
Main Authors:	Xiongjie Jia, Yang Han, Yanjun Li, Yichen Sang, Guolei Zhang
Format:	Article
Language:	English
Published:	Elsevier 2021-11-01
Series:	Energy Reports
Subjects:	Wind turbine Condition monitoring Performance forecasting Denoising autoencoder Residual attention module
Online Access:	http://www.sciencedirect.com/science/article/pii/S2352484721008854

id	doaj-bcca1c468954475eafeaa360eef0db96
record_format	Article
spelling	doaj-bcca1c468954475eafeaa360eef0db962021-10-05T04:19:57ZengElsevierEnergy Reports2352-48472021-11-01763546365Condition monitoring and performance forecasting of wind turbines based on denoising autoencoder and novel convolutional neural networksXiongjie Jia0Yang Han1Yanjun Li2Yichen Sang3Guolei Zhang4College of Power and Energy Engineering, Harbin Engineering University, Harbin, ChinaCollege of Power and Energy Engineering, Harbin Engineering University, Harbin, ChinaCorresponding author.; College of Power and Energy Engineering, Harbin Engineering University, Harbin, ChinaCollege of Power and Energy Engineering, Harbin Engineering University, Harbin, ChinaCollege of Power and Energy Engineering, Harbin Engineering University, Harbin, ChinaWith the proportion of wind power in the grid increasing, the monitoring and maintenance of wind turbines is becoming more and more important for the reliability of the grid. In this study, a data-driven modelling framework based on deep convolutional neural networks is constructed for wind turbines condition monitoring (CM) and performance forecasting (PF). For CM, a robust denoising autoencoder (DAE) model is introduced to output the reconstruction error (RE) of raw signals. The RE is processed to a state indicator by exponentially weighted moving average (EWMA) and monitored on a control chart. For PF, two multi-steps ahead forecasting models are constructed for the forecasting of generator bearing and transformer temperature. To prevent overfitting caused by abundant features, the marginal effect analysis based on random forests is implemented to measure the importance of features. Besides, novel residual attention module (RAM) and training strategies are used improve their representation power of DAE and PF models. Experiments on a real wind turbine dataset prove the effectiveness of the proposed models and methods.http://www.sciencedirect.com/science/article/pii/S2352484721008854Wind turbineCondition monitoringPerformance forecastingDenoising autoencoderResidual attention module
collection	DOAJ
language	English
format	Article
sources	DOAJ
author	Xiongjie Jia Yang Han Yanjun Li Yichen Sang Guolei Zhang
spellingShingle	Xiongjie Jia Yang Han Yanjun Li Yichen Sang Guolei Zhang Condition monitoring and performance forecasting of wind turbines based on denoising autoencoder and novel convolutional neural networks Energy Reports Wind turbine Condition monitoring Performance forecasting Denoising autoencoder Residual attention module
author_facet	Xiongjie Jia Yang Han Yanjun Li Yichen Sang Guolei Zhang
author_sort	Xiongjie Jia
title	Condition monitoring and performance forecasting of wind turbines based on denoising autoencoder and novel convolutional neural networks
title_short	Condition monitoring and performance forecasting of wind turbines based on denoising autoencoder and novel convolutional neural networks
title_full	Condition monitoring and performance forecasting of wind turbines based on denoising autoencoder and novel convolutional neural networks
title_fullStr	Condition monitoring and performance forecasting of wind turbines based on denoising autoencoder and novel convolutional neural networks
title_full_unstemmed	Condition monitoring and performance forecasting of wind turbines based on denoising autoencoder and novel convolutional neural networks
title_sort	condition monitoring and performance forecasting of wind turbines based on denoising autoencoder and novel convolutional neural networks
publisher	Elsevier
series	Energy Reports
issn	2352-4847
publishDate	2021-11-01
description	With the proportion of wind power in the grid increasing, the monitoring and maintenance of wind turbines is becoming more and more important for the reliability of the grid. In this study, a data-driven modelling framework based on deep convolutional neural networks is constructed for wind turbines condition monitoring (CM) and performance forecasting (PF). For CM, a robust denoising autoencoder (DAE) model is introduced to output the reconstruction error (RE) of raw signals. The RE is processed to a state indicator by exponentially weighted moving average (EWMA) and monitored on a control chart. For PF, two multi-steps ahead forecasting models are constructed for the forecasting of generator bearing and transformer temperature. To prevent overfitting caused by abundant features, the marginal effect analysis based on random forests is implemented to measure the importance of features. Besides, novel residual attention module (RAM) and training strategies are used improve their representation power of DAE and PF models. Experiments on a real wind turbine dataset prove the effectiveness of the proposed models and methods.
topic	Wind turbine Condition monitoring Performance forecasting Denoising autoencoder Residual attention module
url	http://www.sciencedirect.com/science/article/pii/S2352484721008854
work_keys_str_mv	AT xiongjiejia conditionmonitoringandperformanceforecastingofwindturbinesbasedondenoisingautoencoderandnovelconvolutionalneuralnetworks AT yanghan conditionmonitoringandperformanceforecastingofwindturbinesbasedondenoisingautoencoderandnovelconvolutionalneuralnetworks AT yanjunli conditionmonitoringandperformanceforecastingofwindturbinesbasedondenoisingautoencoderandnovelconvolutionalneuralnetworks AT yichensang conditionmonitoringandperformanceforecastingofwindturbinesbasedondenoisingautoencoderandnovelconvolutionalneuralnetworks AT guoleizhang conditionmonitoringandperformanceforecastingofwindturbinesbasedondenoisingautoencoderandnovelconvolutionalneuralnetworks
_version_	1716843605580054528

Condition monitoring and performance forecasting of wind turbines based on denoising autoencoder and novel convolutional neural networks

Similar Items