Wind turbine gearbox fault diagnosis based on an improved supervised autoencoder using vibration and motor current signals

Gear box has the high failure rate among the components in wind turbines and the vibration signal analysis is the most common technology in its fault diagnosis. Gearbox vibration signals of wind turbines have complicated compositions, strong backgrou

The problem of the wind turbine gearbox fault diagnosis was investigated by employing the support vector machine (SVM), which is optimized by the improved fruit fly intelligent algorithm with decreasing steps. First of all, the fault characteristic value extracted by Hilbert transform envelope is presented. Secondly, the general SVM solution to the wind turbine gearbox fault diagnosis problem is presented, where the improved fruit fly intelligent algorithm is adopted to optimize the performance of the model of the fault detection and diagnosis. Thirdly, the effectiveness of three fault diagnosis models are compared, including the fault detection and diagnosis model by the traditional SVM, its improved model optimized by the partical swarm optimization algorithm, and improved model by the SVM optimized by the proposed approach. By using the proposed optimization algorithm, the accuracy of gearbox fault diagnosis is much better than other two models, which are validated by the extensive simulation results based on practical historical operation data.

This paper proposes a novel intelligent fault diagnosis method to automatically identify different health conditions of wind turbine (WT) gearbox. Unlike traditional approaches, where feature extraction and classification are separately designed and performed, this paper aims to automatically learn effective fault features directly from raw vibration signals while classify the type of faults in a single framework, thus providing an end-to-end learning-based fault diagnosis system for WT gearbox without additional signal processing and diagnostic expertise. Considering the multiscale characteristics inherent in vibration signals of a gearbox, a new multiscale convolutional neural network (MSCNN) architecture is proposed to perform multiscale feature extraction and classification simultaneously. The proposed MSCNN incorporates multiscale learning into the traditional CNN architecture, which has two merits: 1) high-level fault features can be effectively learned by the hierarchical learning structure with multiple pairs of convolutional and pooling layers; and 2) multiscale learning scheme can capture complementary and rich diagnosis information at different scales. This greatly improves the feature learning ability and enables better diagnosis performance. The proposed MSCNN approach is evaluated through experiments on a WT gearbox test rig. Experimental results and comprehensive comparison analysis with respect to the traditional CNN and traditional multiscale feature extractors have demonstrated the superiority of the proposed method.

Rolling element bearings are very important components in electrical machines. Almost 50% of the faults that occur in the electrical machines occur in the bearings. This makes bearings as one of the most critical components in electrical machinery. Bearing fault diagnosis has drawn the attention of many researchers. Generally, vibration signals from the machine’s accelerometer are used for the diagnosis of bearing faults. In literature, application of Deep Learning algorithms on these vibration signals has resulted in the fault detection accuracy that is close to 100%. Although, fault detection using vibration signals from the machine is ideal but measurement of vibration signals requires an additional sensor, which is absent in many machines, especially low voltage machines as it significantly adds to its cost. Alternatively, bearing fault diagnosis with the help of the stator current or Motor Current Signal (MCS) is also gaining popularity. This paper uses MCS for the diagnosis of bearing inner raceway and outer raceway fault. Diagnosis using MCS is difficult as the fault signatures are buried beneath the noise in the current signal. Hence, signal-processing techniques are employed for the extraction of the fault features. The paper uses the Paderborn University damaged bearing dataset, which contains stator current data from healthy, real damaged inner raceway, and real damaged outer raceway bearings with different fault severity. Fault features are extracted from MCS by first filtering out the redundant frequencies from the signal and then extracting eight features from the filtered signal, which include three features from time domain and five features from time–frequency domain by using the Wavelet Packet Decomposition (WPD). After the extraction of these eight features, the well-known Deep Learning algorithm Long Short-Term Memory (LSTM) is used for bearing fault classification. The Deep Learning LSTM algorithm is mostly used in speech recognition due to its time coherence, but in this paper, the ability of LSTM is also demonstrated with the fault classification accuracy of 97%. A comparison of the proposed algorithm is done with the traditional Machine Learning techniques, and it is shown that the proposed methodology outperforms all the traditional algorithms which are used for the classification of bearing faults using MCS. The method developed is independent of the speed and the loading conditions.

The efficacy of vibration-based fault diagnosis using CNN models could be improved through efficient image extraction from the vibration signals. This study proposed a time-frequency domain vibration image extraction framework based on spectral kurtogram to enhance the classification performance of a Bayesian-tuned CNN model in wind turbine gearbox fault diagnostics. Time-frequency domain kurtograms extracted from vibration signals of a wind turbine gearbox are employed to train a CNN model optimized by Bayesian optimization. Findings of the study show that the kurtogram-based CNN model performed better (recording 171 fewer alarms) than a comparable model trained on images extracted from the time domain vibration signals, highlighting the efficacy of the kurtogram method. In a similar fashion, the Bayesian-optimized CNN model with a classification accuracy of 99.1% recorded 79 fewer false alarms than its standalone counterpart which had an accuracy of 93.2%. It can be concluded that the optimized CNN model trained on Kurtograms is a potential tool for reliable vibration-based condition monitoring of the wind turbine gearbox, minimizing the incidence of false alarms.

Bearing fault diagnosis has extensively exploited vibration signals (VSs) because of their rich information about bearing health conditions. However, this approach is expensive because the measurement of VSs requires external accelerometers. Moreover, in machine systems that are inaccessible or unable to be installed in external sensors, the VS-based approach is impracticable. Otherwise, motor current signals (CSs) are easily measured by the inverters that are the available components of those systems. Therefore, the motor CS-based bearing fault diagnosis approach has attracted considerable attention from researchers. However, the performance of this approach is still not good as the VS-based approach, especially in the case of fault diagnosis for external bearings (the bearings that are installed outside of the electric motors). Accordingly, this article proposes a motor CS-based fault diagnosis method utilizing deep learning and information fusion (IF), which can be applied to external bearings in rotary machine systems. The proposed method uses raw signals from multiple phases of the motor current as direct input, and the features are extracted from the CSs of each phase. Then, each feature set is classified separately by a convolutional neural network (CNN). To enhance the classification accuracy, a novel decision-level IF technique is introduced to fuse information from all of the utilized CNNs. The problem of decision-level IF is transformed into a simple pattern classification task, which can be solved effectively by familiar supervised learning algorithms. The effectiveness of the proposed fault diagnosis method is verified through experiments carried out with actual bearing fault signals.

Deep morphological convolutional network for feature learning of vibration signals and its applications to gearbox fault diagnosis

Studies in condition monitoring literature often aim to detect rolling element bearing faults because they have one of the biggest shares among defects in turbo machinery. Accordingly, several prognosis and diagnosis methods have been devised to identify fault signatures from vibration signals. A recently proposed method to capture the rolling element bearing degradation provides the groundwork for new indicator families utilizing the generalized likelihood ratio test. This novel approach exploits the cyclostationarity and the impulsiveness of vibration signals independently in order to estimate the most suitable indicators for a given fault. However, the method has yet to be tested on complex experimental vibration signals such as those of a wind turbine gearbox. In this study, the approach is applied to the National Renewable Energy Laboratory Wind Turbine Gearbox Condition Monitoring Round Robin Study data set for bearing fault detection purposes. The data set is measured on an experimental test rig of a wind turbine gearbox; hence the complexity of the vibration signals is similar to a real case. The outcome demonstrates that the proposed method is capable of distinguishing between healthy and damaged vibration signals measured on a complex wind turbine gearbox.

Aiming at automated and intelligent state monitoring of induction motors, which are an integral component of a broad spectrum of manufacturing machines, this paper presents a Deep Belief Network (DBN)-based approach to automatically extract relevant features from vibration signals that characterize the working condition of an induction motor. The DBN model employs a structure with stacked restricted Boltzmann machines (RBMs), and is trained by an efficient learning algorithm called greedy layer-wise training. Vibration signals are used as the input to the DBN, and the outputs from activation functions of the trained network are the features needed for fault diagnosis. Comparing to traditional feature extraction methods for induction motor fault diagnosis such as wavelet packet transform, the proposed method is able to learn features directly from the vibration signal to achieve comparable performance with high classification accuracy. Experiments conducted on a machine fault simulator have verified the effectiveness of the proposed method for induction motor fault diagnosis.

Misalignment is one of the most frequent faults observed in rotating machinery. In the present work, the misalignment fault of a motor shaft is studied using multi-scale entropy in combination with a back-propagation neural network algorithm. Experiments were performed, first with an aligned motor shaft, and then with a motor shaft that had angular and parallel misalignment at different operating speeds. Real-time motor current and vibration signals from aligned and different misaligned motor shafts were acquired for the diagnosis of faults. The existing literature mostly focused in the context of frequency-domain analysis. However, in this work multi-scale entropy is used, which accounts for the system complexity. A clear indication of reduction in complexity is observed with the faulty system. The proposed methodology is the first of its kind to detect the misalignment fault using the analysis of both vibration and current signals. The proposed work has also successfully identified for different types of misalignment.

Application of modern signal processing technology, wireless sensor technology, virtual instrument technology, and database technology, the subsystem of remote condition monitoring and fault diagnosis system data analysis and fault diagnosis system is developed. The overall scheme of wind turbine gearbox remote condition monitoring and fault diagnosis system is designed, and the function of each module in the system is introduced. The basic principles of vibration signal noise reduction, time-frequency analysis, and order analysis are described. Through the analysis and diagnosis of measuring vibration signal, the subsystem is verified to be reliable and effective.

One of the main challenges that the industry faces when dealing with massive data for failure diagnosis is high dimensionality of such data. This can be tackled by dimensionality reduction method such as principal components analysis, which usually results in an improved fault diagnosis. Other available techniques include auto-encoders and its variants denoising auto-encoders and sparse auto-encoders. Most recently, variational auto-encoders are one of the most promising techniques for unsupervised learning with successful applications in image processing and speech recognition. Differently from other auto-encoder methods, variational auto-encoders use variational inference to generate a latent representation of the data and impose a distribution over the latent variables and the data itself. In this article, we propose a fully unsupervised deep variational auto-encoder-based approach for dimensionality reduction in fault diagnosis and explore the variational auto-encoder capabilities for such a task. This is achieved by comparing the latent representations provided by variational auto-encoders to the ones from principal components analysis as well as when no reduction is performed in ball bearings’ fault classification using vibration signals. To tackle massive sensor data, we propose different architectures for the variational auto-encoder’s encoder and decoder that are based on deep neural networks and deep convolutional neural networks. Experiments are also carried out by varying the data preprocessing methods for generating spectrograms and hand-engineering features as well as the use of raw vibration data, the architecture of the neural networks fault classifiers operating on the latent representations from variational auto-encoder and principal components analysis, the degree of data dimension reduction, and the size of the available labeled data used for training the fault classifiers. The results show that variational auto-encoders are a competent and promising tool for dimensionality reduction for use in fault diagnosis and worth further exploring their capabilities beyond vibration signals of ball bearing elements.

A Fuzzy Cognitive Network (FCN) is an operational extension of a Fuzzy Cognitive Map (FCM) which incorporates proven stability and guaranteed exponentially-fast error convergence to zero during its training and supports at the same time the continuous interaction with the system it describes. FCN is a suitable, reliable and very powerful FCM structure for system identification, control applications and adaptive decision making because of its convergence based operational architecture. In this paper we introduce the FCN based classifier and apply it on a motor fault detection problem. We use also an alternative form of the network that assumes functional interconnection weights which improves its performance. A benchmark data set is used for condition monitoring of rolling bearings, consisting of motor current and vibration signals. Statistical features are extracted from both current and vibration signals and are combined together to enhance the diagnosis performance of the model. The classification accuracy of the proposed approach shows that the FCN based classifier can be used as a very reliable diagnostic tool for motor fault detection and diagnosis.

Sparsity-enabled signal decomposition using tunable Q-factor wavelet transform for fault feature extraction of gearbox

Variational mode decomposition is a new signal decomposition method, which can process non-linear and non-stationary signals. It can overcome the problems of mode mixing and compensate for the shortcomings in empirical mode decomposition. Permutation entropy is a method which can detect the randomness and kinetic mutation behavior of a time series. It can be considered for use in fault diagnosis. The complexity of wind power generation systems means that the randomness and kinetic mutation behavior of their vibration signals are displayed at different scales. Multi-scale permutation entropy analysis is therefore needed for such vibration signals. This research investigated a method based on variational mode decomposition and permutation entropy for the fault diagnosis of a wind turbine roller bearing. Variational mode decomposition was adopted to decompose the bearing vibration signal into its constituent components. The components containing key fault information were selected for the extraction of their permutation entropy. This entropy was used as a bearing fault characteristic value. The nearest neighbor algorithm was employed as a classifier to identify faults in a roller bearing. The experimental data showed that the proposed method can be applied to wind turbine roller bearing fault diagnosis.