Anomaly detection and fault analysis of wind turbine components based on deep learning network

The usage of machine learning techniques is widely spread and has also been implemented in the wind industry in the last years. Many of these techniques have shown great success but need to constantly prove the expectation of functionality. This paper describes a new method to monitor the health of a wind turbine using an undercomplete autoencoder. To evaluate the health monitoring quality of the autoencoder, the number of anomalies before an event has happened are to be considered. The results show that around 35% of all historical events that have resulted into a failure show many anomalies. Furthermore, the wind turbine subsystems which are subject to good detectability are the rotor system and the control system. If only one third of the service duties can be planned in advance, and thereby the scheduling time can be reduced, huge cost saving potentials can be seen.Energies 2020, 13, 1063 2 of 18 With the model developed, a deviation from normal behavior can be considered as an anomaly. This paper aims to evaluate the anomalies generated by a NBM before a failure has occurred at the WT. These failures are documented in the form of service reports.A literature review is given in Section 2. Hereafter, the NBM is described and trained in Section 3. In Section 3.1, the basic functionality of an Autoencoder (AE) is briefly explained. Data is needed to train the model. This data is described in Section 3.2. Adjustments to the data, filtering and fine-tuning of the parameters of the AE are to be explained in Section 3.3. Thereafter the output of the model as well its usage is presented in Section 3.4.In Section 4, the steps in order to derive a clean failure dataset are shown. Information about downtime events are extracted out of maintenance documents and the measures undertaken at the WT and its components translated into standards (Sections 4.1 and 4.2). To ensure the correctness of the downtime events, we explain how they will be validated in Section 4.4.With the usage of both, the historical failures of the WT (Section 4) and the trained model (Section 3), the time window before the occurrence of a failure can be investigated. This is conducted in Section 5. Subsequently, the detectability of WT failures can be evaluated (Section 5.2). The potential scheduling time and the subsystems with failures are considered to be potentially good detectable in advance. To have a holistic approach, also anomalies during expected normal behavior, are to be investigated. The results are discussed in Section 6 and concluded with Section 7.

show abstract

“…Similar approaches to the AE, as described in this paper, are seen in Zhao et al [11]. Here, findings based on three case studies are explained.…”

Section: Evaluation Of Applicationsupporting

confidence: 63%

“…Approaches with an application of an AE on SCADA data is presented by Zhao et al [11], Vogt et al [12] and Deeskow and Steinmetz [13] among others. Deeskow and Steinmetz [13] have applied ensembles of Autoencoders to detect anomalies in sensor data of various types of power generating systems.…”

Section: Literature Reviewmentioning

confidence: 99%

Evaluation of Anomaly Detection of an Autoencoder Based on Maintenace Information and Scada-Data

et al. 2020

View full text Add to dashboard Cite

show abstract

“…With the development of deep learning technology, the deep neural network (DNN) has obtained great results for renewable energy systems [19,20]. Deep learning is a branch of machine learning which fundamentally aims to use a multilayer neural network to learn the relationship between the input and output in a nonlinear model by mapping the data from the original space to the feature space [21,22]. Compared with traditional ANN methods, DNN can handle more complicated models and is more accurate [23].…”

Section: Introductionmentioning

confidence: 99%

A Novel Condition Monitoring Method of Wind Turbines Based on Long Short-Term Memory Neural Network

et al. 2019

View full text Add to dashboard Cite

Effective intelligent condition monitoring, as an effective technique to enhance the reliability of wind turbines and implement cost-effective maintenance, has been the object of extensive research and development to improve defect detection from supervisory control and data acquisition (SCADA) data, relying on perspective signal processing and statistical algorithms. The development of sophisticated machine learning now allows improvements in defect detection from historic data. This paper proposes a novel condition monitoring method for wind turbines based on Long Short-Term Memory (LSTM) algorithms. LSTM algorithms have the capability of capturing long-term dependencies hidden within a sequence of measurements, which can be exploited to increase the prediction accuracy. LSTM algorithms are therefore suitable for application in many diverse fields. The residual signal obtained by comparing the predicted values from a prediction model and the actual measurements from SCADA data can be used for condition monitoring. The effectiveness of the proposed method is validated in the case study. The proposed method can increase the economic benefits and reliability of wind farms.

show abstract

“…Such discrepancies will reflect whether the machine is in normal or failure mode, which requires a classifier to judge. Recently, some Artificial Intelligent (AI) classifiers, such as neural networks [7][8][9][10][11][12], machine learning methods [13][14][15], and deep learning methods [16,17], have been widely applied in classifying the incipient faults of wind turbines. These methods are really very effective for some faults within a certain working state, but it seems impossible for them to diagnose other faults under other working states.…”

Section: Introductionmentioning

confidence: 99%

Dynamic Fault Monitoring of Pitch System in Wind Turbines using Selective Ensemble Small-World Neural Networks

Wang

2019

Energies

View full text Add to dashboard Cite

Pitch system failures occur primarily because wind turbines typically work in dynamic and variable environments. Conventional monitoring strategies show limitations of continuously identifying faults in most cases, especially when rapidly changing winds occur. A novel selective-ensemble monitoring strategy is presented to diagnose the most pitch failures using Supervisory Control and Data Acquisition (SCADA) data. The proposed strategy consists of five steps. During the first step, the SCADA data are partitioned according to the turbine’s four working states. Correlation Information Entropy (CIE) and 10 indicators are used to select correlation signals and extract features of the partition data, respectively. During the second step, multiple Small-World Neural Networks (SWNNs) are established as the ensemble members. Regarding the third step, all the features are randomly sampled to train the SWNN members. The fourth step involves using an improved global correlation method to select appropriate ensemble members while in the fifth step, the selected members are fused to obtain the final classification result based on the weighted integration approach. Compared with the conventional methods, the proposed ensemble strategy shows an effective accuracy rate of over 93.8% within a short delay time.

show abstract

Anomaly detection and fault analysis of wind turbine components based on deep learning network

Cited by 247 publications

References 19 publications

Evaluation of Anomaly Detection of an Autoencoder Based on Maintenace Information and Scada-Data

Evaluation of Anomaly Detection of an Autoencoder Based on Maintenace Information and Scada-Data

A Novel Condition Monitoring Method of Wind Turbines Based on Long Short-Term Memory Neural Network

Dynamic Fault Monitoring of Pitch System in Wind Turbines using Selective Ensemble Small-World Neural Networks

Contact Info

Product

Resources

About