Search

Wind turbines (WTs) generally comprise several complex and interconnected systems, such as hub, converter, gearbox, generator, yaw system, pitch system, hydraulic system control system,integration control system, and auxiliary system. Moreover, fault diagnosis plays an important role in ensuring WT safety. In the past decades, machine learning (ML) has showed a powerful capability in fault detection and diagnosis of WTs, thereby remarkably reducing equipment downtime and minimizing financial losses. This study provides a comprehensive review of recent studies on ML methods and techniques for WT fault diagnosis. These studies are classified as supervised, unsupervised, and semi-supervised learning methods. Existing state-of-the-art methods are analyzed and characteristics are discussed. Perspectives on challenges and further directions are also provided.

Aiming at the problem of difficulties in modeling the nonlinear relation in the steam coal dataset, this article proposes a forecasting method for the price of steam coal based on robust regularized kernel regression and empirical mode decomposition. By selecting the polynomial kernel function, the robust loss function and L2 regular term to construct a robust regularized kernel regression model are used. The polynomial kernel function does not depend on the kernel parameters and can mine the global rules in the dataset so that improves the forecasting stability of the kernel model. This method maps the features to the high-dimensional space by using the polynomial kernel function to transform the nonlinear law in the original feature space into linear law in the high-dimensional space and helps learn the linear law in the high-dimensional feature space by using the linear model. The Huber loss function is selected to reduce the influence of abnormal noise in the dataset on the model performance, and the L2 regular term is used to reduce the risk of model overfitting. We use the combined model based on empirical mode decomposition (EMD) and auto regressive integrated moving average (ARIMA) model to compensate for the error of robust regularized kernel regression model, thus making up for the limitations of the single forecasting model. Finally, we use the steam coal dataset to verify the proposed model and such model has an optimal evaluation index value compared to other contrast models after the model performance is evaluated as per the evaluation index such as RMSE, MAE, and mean absolute percentage error.

A fault detection method of wind turbine pitch system using semi-supervised optimal margin distribution machine (ssODM) optimized by dynamic state transition algorithm (DSTA) [ssODM-DSTA] was proposed to solve the problem of obtaining the optimal hyperparameters of the fault detection model for the pitch system. This method was adopted to input the three hyperparameters of the ssODM into the dynamic state transition algorithm in the form of a three-dimensional vector to obtain the global optimal hyperparameters of the model, thus improving the performance of the fault detection model. Using a random forest to rank the priority of features of the pitch system fault data, the features with large weight proportions were retained. Then, the Pearson correlation method is used to analyze the degree of correlation among features, filter redundant features, and reduce the scale of features. The dataset was divided into a training dataset and a test dataset to train and test the proposed fault detection model, respectively. The real-time wind turbine pitch system fault data were collected from domestic wind farms to carry out fault detection experiments. The results have shown that the proposed method had a positive fault rate (FPR) and fault negative rate (FNR), compared with other optimization algorithms.

Aiming at the problem of unbalanced data categories of UHV converter valve fault data, a method for UHV converter valve fault detection based on optimization cost-sensitive extreme random forest is proposed. The misclassification cost gain is integrated into the extreme random forest decision tree as a splitting index, and the inertia weight and learning factor are improved to construct an improved particle swarm optimization algorithm. First, feature extraction and data cleaning are carried out to solve the problems of local data loss, large computational load, and low real-time performance of the model. Then, the classifier training based on the optimization cost-sensitive extreme random forest is used to construct a fault detection model, and the improved particle swarm optimization algorithm is used to output the optimal model parameters, achieving fast response of the model and high classification accuracy, good robustness, and generalization under unbalanced data. Finally, in order to verify its effectiveness, this model is compared with the existing optimization algorithms. The running speed is faster and the fault detection performance is higher, which can meet the actual needs.

In practice, faulty samples of wind turbine (WT) gearboxes are far smaller than normal samples during operation, and most of the existing fault diagnosis methods for WT gearboxes only focus on the improvement of classification accuracy and ignore the decrease of missed alarms and the reduction of the average cost. To this end, a new framework is proposed through combining the Spearman rank correlation feature extraction and cost-sensitive LightGBM algorithm for WT gearbox’s fault detection. In this article, features from wind turbine supervisory control and data acquisition (SCADA) systems are firstly extracted. Then, the feature selection is employed by using the expert experience and Spearman rank correlation coefficient to analyze the correlation between the big data of WT gearboxes. Moreover, the cost-sensitive LightGBM fault detection framework is established by optimizing the misclassification cost. The false alarm rate and the missed detection rate of the WT gearbox under different working conditions are finally obtained. Experiments have verified that the proposed method can significantly improve the fault detection accuracy. Meanwhile, the proposed method can consistently outperform traditional classifiers such as AdaCost, cost-sensitive GBDT, and cost-sensitive XGBoost in terms of low false alarm rate and missed detection rate. Owing to its high Matthews correlation coefficient scores and low average misclassification cost, the cost-sensitive LightGBM (CS LightGBM) method is preferred for imbalanced WT gearbox fault detection in practice.

Aiming at the problem of class imbalance in the wind turbine blade bolts operation-monitoring dataset, a fault detection method for wind turbine blade bolts based on Gaussian Mixture Model–Synthetic Minority Oversampling Technique–Gaussian Mixture Model (GSG) combined with Cost-Sensitive LightGBM (CS-LightGBM) was proposed. Since it is difficult to obtain the fault samples of blade bolts, the GSG oversampling method was constructed to increase the fault samples in the blade bolt dataset. The method obtains the optimal number of clusters through the BIC criterion, and uses the GMM based on the optimal number of clusters to optimally cluster the fault samples in the blade bolt dataset. According to the density distribution of fault samples in inter-clusters, we synthesized new fault samples using SMOTE in an intra-cluster. This retains the distribution characteristics of the original fault class samples. Then, we used the GMM with the same initial cluster center to cluster the fault class samples that were added to new samples, and removed the synthetic fault class samples that were not clustered into the corresponding clusters. Finally, the synthetic data training set was used to train the CS-LightGBM fault detection model. Additionally, the hyperparameters of CS-LightGBM were optimized by the Bayesian optimization algorithm to obtain the optimal CS-LightGBM fault detection model. The experimental results show that compared with six models including SMOTE-LightGBM, CS-LightGBM, K-means-SMOTE-LightGBM, etc., the proposed fault detection model is superior to the other comparison methods in the false alarm rate, missing alarm rate and F1-score index. The method can well realize the fault detection of large wind turbine blade bolts.

As one of the key components of wind turbines, gearboxes are under complex alternating loads for a long time, and the safety and reliability of the whole machine are often affected by the failure of internal gears and bearings. Aiming at the difficulty of optimizing the parameters of wind turbine gearbox fault detection models based on extreme random forest, a fault detection model with extreme random forest optimized by the improved butterfly optimization algorithm (IBOA-ERF) is proposed. The algebraic sum of the false alarm rate and the missing alarm rate of the fault detection model is constructed as the fitness function, and the initial position and position update strategy of the individual are improved. A chaotic mapping strategy is introduced to replace the original population initialization method to enhance the randomness of the initial population distribution. An adaptive inertia weight factor is proposed, combined with the landmark operator of the pigeon swarm optimization algorithm to update the population position iteration equation to speed up the convergence speed and improve the diversity and robustness of the butterfly optimization algorithm. The dynamic switching method of local and global search stages is adopted to achieve dynamic balance between global exploration and local search, and to avoid falling into local optima. The ERF fault detection model is trained, and the improved butterfly optimization algorithm is used to obtain optimal parameters to achieve fast response of the proposed model with good robustness and generalization under high-dimensional data. The experimental results show that, compared with other optimization algorithms, the proposed fault detection method of wind turbine gearboxes has a lower false alarm rate and missing alarm rate.

The number of normal samples of wind turbine generators is much larger than the number of fault samples. To solve the problem of imbalanced classification in wind turbine generator fault detection, a cost-sensitive extremely randomized trees (CS-ERT) algorithm is proposed in this paper, in which the cost-sensitive learning method is introduced into an extremely randomized trees (ERT) algorithm. Based on the classification misclassification cost and class distribution, the misclassification cost gain (MCG) is proposed as the score measure of the CS-ERT model growth process to improve the classification accuracy of minority classes. The Hilbert-Schmidt independence criterion lasso (HSICLasso) feature selection method is used to select strongly correlated non-redundant features of doubly-fed wind turbine generators. The effectiveness of the method was verified by experiments on four different failure datasets of wind turbine generators. The experiment results show that average missing detection rate, average misclassification cost and gMean of the improved algorithm better than those of the ERT algorithm. In addition, compared with the CSForest, AdaCost and MetaCost methods, the proposed method has better real-time fault detection performance.

A fault diagnosis method for wind turbine gearboxes based on undersampling, XGBoost feature selection, and improved whale optimization-random forest (IWOA-RF) was proposed for the problem of high false negative and false positive rates in wind turbine gearboxes. Normal samples of raw data were subjected to undersampling first, and various features and data labels in the raw data were provided with importance analysis by XGBoost feature selection to select features with higher label correlation. Two parameters of random forest algorithm were optimized via the whale optimization algorithm to create a fitness function with the false negative rate (FNR) and false positive rate (FPR) as evaluation indexes. Then, the minimum fitness function value within the given scope of parameters was found. The WOA was controlled by the hyper-parameter α to optimize the step size. This article uses the variant form of the sigmoid function to alter the change trend of the WOA hyper-parameter α from a linear decline to a rapid decline first and then a slow decline to allow the WOA to be optimized. In the initial stage, a larger step size and step size change rate can make the model progress to the optimization target faster, while in the later stage of optimization, a smaller step size and step size change rate allows the model to more accurately find the minimum value of the fitness function. Finally, two hyper-parameters, corresponding to the minimum fitness function value, were substituted into a random forest algorithm for model training. The results showed that the method proposed in this paper can significantly reduce the false negative and false positive rates compared with other optimization classification methods.

It is difficult to optimize the fault model parameters when Extreme Random Forest is used to detect the electric pitch system fault model of the double-fed wind turbine generator set. Therefore, Extreme Random Forest which was optimized by improved grey wolf algorithm (IGWO-ERF) was proposed to solve the problems mentioned above. First, IGWO-ERF imports the Cosine model to nonlinearize the linearly changing convergence factor α to balance the global exploration and local exploitation capabilities of the algorithm. Then, in the later stage of the algorithm iteration, α wolf generates its mirror wolf based on the lens imaging learning strategy to increase the diversity of the population and prevent local optimum of the population. The electric pitch system fault detection method of the wind turbine generator set sets the generator power of the variable pitch system as the main state parameter. First, it uses the Pearson correlation coefficient method to eliminate the features with low correlation with the electric pitch system generator power. Then, the remaining features are ranked by the importance of the RF features. Finally, the top N features are selected to construct the electric pitch system fault data set. The data set is divided into a training set and a test set. The training set is used to train the proposed fault detection model, and the test set is used for testing. Compared with other parameter optimization algorithms, the proposed method has lower FNR and FPR in the electric pitch system fault detection of the wind turbine generator set.

To address the issue of a large calculation and difficult optimization for the traditional fault detection of a wind turbine-based pitch control system, a fault detection model, based on LightGBM by the improved Harris Hawks optimization algorithm (light gradient boosting machine by the improved Harris Hawks optimization, IHHO-LightGBM) for the wind turbine-based pitch control system, is proposed in this article. Firstly, a trigonometric function model is introduced by IHHO to update the prey escape energy, to balance the global exploration ability and local development ability of the algorithm. In this model, the fault detection false alarm rate is used as the fitness function, and the two parameters are used as the optimization objects of the improved Harris Hawks optimization algorithm, to optimize the parameters, so as to achieve the global optimal parameters to improve the performance of the fault detection model. Three different fault data of the pitch control system in actual operations of domestic wind farms are used as the experimental data, the Pearson correlation analysis method is introduced, and the wind turbine power output is taken as the main state parameter, to analyze the correlation degree of all the characteristic variables of the data and screen the important characteristic variables out, so as to achieve the effective dimensionality reduction process of the data, by using the feature selection method. Three established fault detection models are selected and compared with the proposed method, to verify its feasibility. The experimental data indicate that compared with other algorithms, the fault detecting ability of the proposed model is improved in all aspects, and the false alarm rate and false negative rate are lower.

It is widely accepted that conventional boost algorithms are of low efficiency and accuracy in dealing with big data collected from wind turbine operations. To address this issue, this paper is devoted to the application of an adaptive LightGBM method for wind turbine fault detections. To this end, the realization of feature selection for fault detection is firstly achieved by utilizing the maximum information coefficient to analyze the correlation among features in supervisory control and data acquisition (SCADA) of wind turbines. After that, a performance evaluation criterion is proposed for the improved LightGBM model to support fault detections. In this scheme, by embedding the confusion matrix as a performance indicator, an improved LightGBM fault detection approach is then developed. Based on the adaptive LightGBM fault detection model, a fault detection strategy for wind turbine gearboxes is investigated. To demonstrate the applications of the proposed algorithms and methods, a case study with a three-year SCADA dataset obtained from a wind farm sited in Southern China is conducted. Results indicate that the proposed approaches established a fault detection framework of wind turbine systems with either lower false alarm rate or lower missing detection rate.

Artificial bee colony (ABC) is a novel population-based optimization method, having the advantage of less control parameters, being easy to implement, and having strong global optimization ability. However, ABC algorithm has some shortcomings concerning its position-updated equation, which is skilled in global search and bad at local search. In order to coordinate the ability of global and local search, we first propose a self-adaptive ABC algorithm (denoted as SABC) in which an improved position-updated equation is used to guide the search of new candidate individuals. In addition, good-point-set approach is introduced to produce the initial population and scout bees. The proposed SABC is tested on 12 well-known problems. The simulation results demonstrate that the proposed SABC algorithm has better search ability with other several ABC variants.

Fault diagnosis and forecasting contribute significantly to the reduction of operating and maintenance associated costs, as well as to improve the resilience of wind turbine systems. Different from the existing fault diagnosis approaches using monitored vibration and acoustic data from the auxiliary equipment, this research presents a novel fault diagnosis and forecasting approach underpinned by a support vector regression model using data obtained by the supervisory control and data acquisition system (SCADA) of wind turbines (WT). To operate, the extraction of fault diagnosis features is conducted by measuring SCADA parameters. After that, confidence intervals are set up to guide the fault diagnosis implemented by the support vector regression (SVR) model. With the employment of confidence intervals as the performance indicators, an SVR-based fault detecting approach is then developed. Based on the WT SCADA data and the SVR model, a fault diagnosis strategy for large-scale doubly-fed wind turbine systems is investigated. A case study including a one-year monitoring SCADA data collected from a wind farm in Southern China is employed to validate the proposed methodology and demonstrate how it works. Results indicate that the proposed strategy can support the troubleshooting of wind turbine systems with high precision and effective response.

Cost-sensitive support vector machine is one of the most popular tools to deal with class-imbalanced problem such as fault diagnosis. However, such data appear with a huge number of examples as well as features. Aiming at class-imbalanced problem on big data, a cost-sensitive support vector machine using randomized dual coordinate descent method (CSVM-RDCD) is proposed in this paper. The solution of concerned subproblem at each iteration is derived in closed form and the computational cost is decreased through the accelerating strategy and cheap computation. The four constrained conditions of CSVM-RDCD are derived. Experimental results illustrate that the proposed method increases recognition rates of positive class and reduces average misclassification costs on real big class-imbalanced data.