Generative Transfer Learning for Intelligent Fault Diagnosis of the Wind Turbine Gearbox

Guo, Jianwen; Wu, Jiapeng; Zhang, Shaohui; Long, Jianyu; Chen, W.; Cabrera, Diego; Li, Chuan

doi:10.3390/s20051361

Cited by 29 publications

(16 citation statements)

References 29 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In the experiment, the rotation speed of the generator was 1590rpm. According to the experimental standard of wind turbine [30], the fault signal of the testing bearing was collected by a accelerometer, the sampling frequency was 12.8kHz, as shown in Table 2. There are four fault types and 400 samples in the data set D, and each sample has 1200 sampling points.…”

Section: A Data Descriptionmentioning

confidence: 99%

Deep Transfer Network With Multi-Kernel Dynamic Distribution Adaptation for Cross-Machine Fault Diagnosis

Liu

et al. 2021

IEEE Access

View full text Add to dashboard Cite

Recently, various deep learning models, which are mainly based on data-driven algorithms, have received more and more attention in the field of intelligent fault diagnosis and prognostics. However, there are two major assumptions accepted by default in the existing studies: 1) The training (source domain) and testing (target domain) data sets obey the same feature distribution; 2) Sufficient labeled data with fault information is available for model training. In real industrial scenarios, especially for different machines, these assumptions are mostly invalid, which makes it a huge challenge to build reliable diagnostic model. Motivated by transfer learning, we present a novel intelligent method named deep transfer network (DTN) with multi-kernel dynamic distribution adaptation (MDDA) to address the problem of cross-machine fault diagnosis. In the proposed approach, the DTN has wide first-layer convolutional kernel and several small convolutional layers, which is utilized to extract transferable features across different machines and suppress high frequency noise. Then, the MDDA method constructs a weighted mixed kernel function to map different transferable features to a unified feature space, and the relative importance of the marginal and conditional distributions are also evaluated dynamically. The proposed method is verified by three transfer learning tasks of bearings, in which the health states of wind turbine bearings in real scenario are identified by using diagnosis knowledge from two different bearings in laboratories. The results show that the proposed method can achieve higher diagnosis accuracy and better transfer performance even under different noisy environment conditions than many other state-of-the-art methods. The presented framework offers a promising approach for cross-machine fault diagnosis. INDEX TERMS Deep transfer network, multi-kernel dynamic distribution adaptation, cross-machine fault diagnosis, transfer learning, bearings.

show abstract

Section: A Data Descriptionmentioning

confidence: 99%

Deep Transfer Network With Multi-Kernel Dynamic Distribution Adaptation for Cross-Machine Fault Diagnosis

Liu

et al. 2021

IEEE Access

View full text Add to dashboard Cite

show abstract

“…In the limit switch failure case, the SCADA alarm did not detect it, but the technique did before failure. Guo et al [24] presented a method of normal behaviour modelling for pitch system failure detection. The authors used a multivariate Gaussian process to predict power output and it was found to have smaller residuals and MAE compared to both the binned power curve and sixth-order polynomial model.…”

Section: Pitch System Condition Monitoringmentioning

confidence: 99%

Investigation of Isolation Forest for Wind Turbine Pitch System Condition Monitoring Using SCADA Data

et al. 2021

View full text Add to dashboard Cite

Wind turbine pitch system condition monitoring is an active area of research, and this paper investigates the use of the Isolation Forest Machine Learning model and Supervisory Control and Data Acquisition system data for this task. This paper examines two case studies, turbines with hydraulic or electric pitch systems, and uses an Isolation Forest to predict failure ahead of time. This novel technique compared several models per turbine, each trained on a different number of months of data. An anomaly proportion for three different time-series window lengths was compared, to observe trends and peaks before failure. The two cases were compared, and it was found that this technique could detect abnormal activity roughly 12 to 18 months before failure for both the hydraulic and electric pitch systems for all unhealthy turbines, and a trend upwards in anomalies could be found in the immediate run up to failure. These peaks in anomalous behaviour could indicate a future failure and this would allow for on-site maintenance to be scheduled. Therefore, this method could improve scheduling planned maintenance activity for pitch systems, regardless of the pitch system employed.

show abstract

“…Wen et al [8] also developed a deep transfer learning method for fault diagnosis that was tested on bearing data sets collected under different loading conditions. In addition, to solve the problem of the significantly degraded performance of the traditional intelligent fault diagnosis algorithm degrades for changes in the workload, Guo et al [9] proposed a transfer learning method and verified it by experimenting on the fault diagnosis of the wind turbine gearbox.…”

Section: Introductionmentioning

confidence: 99%

A Deep Transfer Learning Method for Bearing Fault Diagnosis Based on Domain Separation and Adversarial Learning

Xiang

Zhang

Gao

et al. 2021

Shock and Vibration

View full text Add to dashboard Cite

Current studies on intelligent bearing fault diagnosis based on transfer learning have been fruitful. However, these methods mainly focus on transfer fault diagnosis of bearings under different working conditions. In engineering practice, it is often difficult or even impossible to obtain a large amount of labeled data from some machines, and an intelligent diagnostic method trained by labeled data from one machine may not be able to classify unlabeled data from other machines, strongly hindering the application of these intelligent diagnostic methods in certain industries. In this study, a deep transfer learning method for bearing fault diagnosis, domain separation reconstruction adversarial networks (DSRAN), was proposed for the transfer fault diagnosis between machines. In DSRAN, domain-difference and domain-invariant feature extractors are used to extract and separate domain-difference and domain-invariant features, respectively Moreover, the idea of generative adversarial networks (GAN) was used to improve the network in learning domain-invariant features. By using domain-invariant features, DSRAN can adopt the distribution of the data in the source and target domains. Six transfer fault diagnosis experiments were performed to verify the effectiveness of the proposed method, and the average accuracy reached 89.68%. The results showed that the DSRAN method trained by labeled data obtained from one machine can be used to identify the health state of the unlabeled data obtained from other machines.

show abstract

Generative Transfer Learning for Intelligent Fault Diagnosis of the Wind Turbine Gearbox

Cited by 29 publications

References 29 publications

Deep Transfer Network With Multi-Kernel Dynamic Distribution Adaptation for Cross-Machine Fault Diagnosis

Deep Transfer Network With Multi-Kernel Dynamic Distribution Adaptation for Cross-Machine Fault Diagnosis

Investigation of Isolation Forest for Wind Turbine Pitch System Condition Monitoring Using SCADA Data

A Deep Transfer Learning Method for Bearing Fault Diagnosis Based on Domain Separation and Adversarial Learning

Contact Info

Product

Resources

About