Gearbox Fault Diagnosis Research Articles

Fault Diagnosis of Gearbox Based on CNN and LSTM with Attention Mechanism

Gearbox is a key component of mechanical equipment, which has a complex structure, harsh working conditions, and a higher probability of failure. Therefore, gearbox fault diagnosis is vital to ensure the efficient operation of the whole mechanical equipment. However, traditional gearbox fault diagnosis methods mainly rely on manual feature extraction. To address this issue, a novel end-to-end fault diagnosis model that combines convolutional neural network (CNN), long short-term memory network (LSTM) and attention mechanism (AM) is proposed. Firstly, Spatial features are extracted from the original input data using CNNs, and then temporal features are extracted even further from the spatial features using LSTMs. The attention mechanism improves the network's attention to the global key features by assigning weights, and finally, the fault diagnosis results are derived from the fully connected layer and Softmax classifier. The experimental results show that the method is able to adaptively extract features and ensure higher diagnostic accuracy, validating the effectiveness of the proposed method.

Study on the Vibration Characteristics of the Helical Gear-Rotor-Bearing Coupling System of a Wind Turbine with Composite Faults

As the core component of the wind turbine generation gearbox, the gear-rotor-bearing transmission system typically operates in harsh environments, inevitably leading to the occurrence of composite faults in the system, which exacerbates system vibration. Therefore, it is necessary to study the vibration characteristics of wind turbine helical gear-rotor-bearing transmission systems with composite faults. This paper uses an improved energy method to calculate the theoretical time-varying mesh stiffness of a helical gear with a root crack failure. On the premise of considering the time-varying meshing stiffness of the faulty helical gear, the gear eccentric fault, and the nonlinear support force of the faulty bearing, a multi-degree-of-freedom helical gear-rotor-bearing transmission system with compound faults was established by using the lumped parameter method. The dynamic model of the system was solved based on the Runge–Kutta method, and the vibration response of the system under healthy conditions, single faults with gear eccentricity, single faults with tooth root cracks, and coupled bearing composite faults were simulated and analyzed. The results show that the simulation results based on KISSsoft software 2018 version verify the effectiveness of the improved energy method; the existence of single faults and composite faults will cause the fault characteristics in the time domain and frequency domain responses. In this paper, the influence of a single fault and a complex fault on the time domain and frequency domain of the system is mainly discovered through the fault study of the helical rotor-bearing system, and the influence of the fault degree on the vibration of the gear motion system is discussed. The greater the degree of the fault, the more vibration of the system occurs; accordingly, when the system is under the coupling of tooth root crack and bearing fault, there is a significant difference compared with the healthy system and the single fault system. The system vibration has obvious time domain and frequency domain signal characteristics, including periodic pulse impacts caused by gear faults and time domain impact caused by bearing. The fault characteristic frequencies can also be found in the frequency domain. In this paper, the fault study of a helical gear of wind turbine generation provides a reference for the theoretical analysis of the vibration characteristics of the helical gear-rotor-bearing system under various fault conditions, lays a solid foundation for the simulation and subsequent diagnosis of the composite fault signal of the system, and provides help for the fault diagnosis of wind turbine gearboxes in the future.

Detect Gear Fault in Wind Turbine Gearbox based on Time Synchornous Averaging Without Phase Signal.

Wind energy is increasingly recognized as a worldwide sustainable and environmentally friendly power source. Nevertheless, an important barrier to further investment in wind energy is the large rate of failure that occurs to turbines. The gearbox, a crucial component, has a major effect on the performance of wind turbines. It consists of complex planetary and cylindrical gear systems, making it prone to failure and causing major defects in wind turbines. Therefore, there is an urgent need to minimize downtime and improve productivity in wind turbine gearbox operations. Recent decades have witnessed growing interest in fault diagnosis of gearboxes due to their widespread use and industry significance. The time synchronous averaging (TSA) method is widely used as a fundamental approach to detect faults in wind turbine gearboxes from mechanical vibration signals. Usually, applying this method requires a device to measure the vibration phase. Nevertheless, there are certain circumstances where installing a phase-measuring device might pose challenges. For instance, if the gearbox operates, it cannot be paused for installation. Additionally, if the gearboxes are enclosed, it becomes difficult to insert the device. The present paper presents an innovative technical approach to improve the time synchornous averaging method without requiring information about phase vibration. It also evaluates the effectiveness of the proposed method using feature values. An experimental test rig was set up to evaluate the algorithm's effectiveness, simulating different gear faults and load conditions.

Investigation on the fusion reliability and cluster consistency of multivariable entropy method

Recent researches have shown that the multivariable entropy based feature extraction method can obtain better diagnosis results for fault diagnosis of planetary gearbox. However, the nature properties of multivariable entropy have still not been deeply explored: the reliability of multi-source information fusion and cluster consistency for the same fault signal. These two properties will affect the accuracy of fault diagnosis based on multivariate entropy. This paper aims to reveal the nature properties of multivariate entropy. Firstly, a rigid-flexible coupling dynamic model of a planetary gearbox is conducted to establish a pure test environment. Then the generated vibration signals are used to evaluate the fusion reliability and cluster consistency of multivariable entropy. Additionally, a new multivariable entropy feature extraction method called variational embedding refined composite multiscale diversity entropy (veRCMDE) is proposed. Finally, the simulation and experiment results show that high fusion reliability and high cluster consistency enable multivariate entropy to extract more valuable features, and the proposed veRCMDE performs the best in all experiments.

Application of On-Rotor Sensing in Planetary Gearbox Fault Diagnosis

Planetary gearbox is a crucial mechanical transmission component, widely used in various mechanical equipment and industrial fields. Its failure will not only cause significant economic losses, but also may bring threats to the life and health of workers. Therefore, the condition monitoring and fault diagnosis of planetary gearbox is of great significance. In this paper, a novel on-rotor sensing (ORS) technique based on a three-axis MEMS accelerometer is employed to collect vibration signals from planetary gearboxes, in which the accelerometer is mounted on the shaft near to the sun gear and rotates together with the shaft. The two orthogonal outputs of the collected vibration signals is used to construct an analytic signal and then demodulated to get the envelope spectrum, which is then used diagnose the fault in the planetary gearbox. Results show that the proposed method can effectively diagnose the sun gear crack fault, while conventional piezoelectric sensors placed on planetary gearbox casing are less capable of detecting such faults, demonstrating the superiority of ORS technology in the field of rotating machinery fault diagnosis.

Optimal weight impulse extraction: New impulse extraction methodology for incipient gearbox condition monitoring

Gear faults in a transmission system generally cause impulse components in vibration signals, which is a crucial symbol for gearbox fault diagnosis. However, their related signals are often interfered or even submerged by the noisy meshing components (NMC) of gearboxes in degradation, which introduces challenges for incipient fault detection and condition monitoring. Commonly employed deconvolution-based methods attempt to design a filter to extract impulse components. However, these methods fail to address the interference issue of the NMC on deconvolution process. To overcome this limitation, this paper proposes an optimal weight impulse extraction (OWIE) methodology to suppress the NMC and highlight impulse component in vibration signal. Different from deconvolution-based methods, the proposed method locates impulses adaptively driven by waveform itself. A non-impulse part is suppressed through point-to-point removal, while impulse components are highlighted by subtracting a weighted signal from a raw signal. An iterative procedure is utilized to solve the optimal weight sequence by maximizing the kurtosis of an impulse signal. The effectiveness of the proposed OWIE is validated through a simulation case study and a run-to-failure experiment of gearboxes. Results demonstrate that the OWIE is sensitive to incipient faults and is suitable for the health condition monitoring of gearboxes.

Comparative Analysis of SVM and ANN for Machine Condition Monitoring and Fault Diagnosis in Gearboxes

Comparative Analysis of SVM and ANN for Machine Condition Monitoring and Fault Diagnosis in Gearboxes

A comparative study on ML algorithms and ANN on fault diagnosis of a two-stage helical gearbox

In the modern world, condition monitoring is crucial to the predictive maintenance of machinery. Gearboxes are widely used in machineries and auto motives to achieve the variable speeds. The major problem in gearbox is catastrophic failure due to heavy loads, corrosion and erosion, results in economic loss and creates high safety risks. So, it is necessary to provide condition monitoring technique to detect and diagnose failures, to achieve cost benefits to industry. The main purpose of this study to use Machine Learning (ML) algorithms and Artificial Neural Network (ANN) which are very powerful and reliable tool for fault detection and its most important attribute is its ability to efficiently detect non-stationary, non-periodic, transient features of the vibration signal. To do the vibration study, an experimental setup was created, and various faults were induced faults of various kinds that usually occurred in the gearbox. The gear in the gear train was subjected to vibration analysis which was captured via a sensor. Signal processing was carried out using MATLAB Toolbox. To automatically identify the flaws in the helical gearbox, an artificial neural network (ANN) and several machines learning methods, including KNN, decision tree, random forest, and SMV, were trained by creating a database from the experiment conducted. The outcomes showed potential in accurately classifying the faults. The results show that ANN has the highest accuracy of 99.6% with a 6.5662 seconds computational time while SVM has the lowest accuracy of 96% among them along with the highest computational time of 21.324 seconds.

Enhancing Gearbox Fault Diagnosis through Advanced Feature Engineering and Data Segmentation Techniques

Efficient gearbox fault diagnosis is crucial for the cost-effective maintenance and reliable operation of rotating machinery. Despite extensive research, effective fault diagnosis remains challenging due to the multitude of features available for classification. Traditional feature selection methods often fail to achieve optimal performance in fault classification tasks. This study introduces diverse ranking methods for selecting the relevant features and utilizes data segmentation techniques such as sliding, windowing, and bootstrapping to strengthen predictive model performance and scalability. A comparative analysis of these methods was conducted to identify the potential causes and future solutions. An evaluation of the impact of enhanced feature engineering and data segmentation on predictive maintenance in gearboxes revealed promising outcomes, with decision trees, SVM, and KNN models outperforming others. Additionally, within a fully connected network, windowing emerged as a more robust and efficient segmentation method compared to bootstrapping. Further research is necessary to assess the performance of these techniques across diverse datasets and applications, offering comprehensive insights for future studies in fault diagnosis and predictive maintenance.

Fault diagnosis of gearboxin wind turbine based on EMD-DCGAN

INTRODUCTION: Wind turbine gearbox fault diagnosis is of great significance for the safe and stable operation of wind turbines. The accuracy of wind turbine gearbox fault diagnosis can be effectively improved by using complete wind turbine gearbox fault data and efficient fault diagnosis algorithms.A wind turbine gearbox fault diagnosis method based on EMD-DCGAN method is proposed in this paper.
 OBJECTIVES: It can solve the problem when the sensor fails or the data transmission fails, it will lead to errors in the wind turbine gearbox fault data, which in turn will lead to a decrease in the wind turbine gearbox fault diagnosis accuracy.
 METHODS: Firstly, the outliers in the sample data need to be detected and removed. In this paper, the EMD method is used to eliminate outliers in the wind turbine gearbox fault data samples with the aim of enhancing the true continuity of the samples; secondly, in order to make up for the lack of missing samples, a data enhancement algorithm based on a GAN network is proposed in the paper, which is able to effectively perfect the missing items of the sample data; lastly, in order to improve the accuracy of wind turbine gearbox faults, a DCGAN neural network-based fault diagnosis method is proposed, which effectively combines the data dimensionality reduction feature of deep learning method and the data enhancement feature of generative adversarial network, and can improve the accuracy and speed of fault diagnosis.
 RESULTS and CONCLUSIONS: The experimental results show that the proposed method can effectively identify wind turbine gearbox fault conditions, and verify the effectiveness of the algorithm under different sample data conditions.

Local maximum synchrosqueezing reassigning chirplet transform and its application to gearbox fault diagnosis

Abstract Aiming at the problems of poor time-frequency aggregation and severe noise interference when traditional time-frequency analysis methods deal with complex multi-component signals, this paper proposes a new time-frequency analysis method – local maximum synchrosqueezing reassigning chirplet transform (LMSRCT). The core idea of the method is to introduce the principle of general linear chirplet transform (GLCT) into synchro-reassigning transform (SRT), followed by reassigning the results of local maximum synchrosqueezing chirplet transform (LMSSCT), and then introducing the innovative synchro-reassigning operator (SRO). This results to a novel three-step method for extracting instantaneous frequency, which ultimately yields the final time-frequency representation. This method ensures the integrity of each instantaneous frequency, solves the problem of energy divergence problems, and improves aggregation. Simulation and experimental results show that the LMSRCT can provide better characterization results compared to other methods and can effectively solve the different situations that occur between the instantaneous frequencies of complex signals. The proposed method can effectively estimate transmission speed, which achieves fault diagnosis under tacholess conditions, and the order analysis results are more accurate and reliable.

A gradient aligned domain adversarial network for unsupervised intelligent fault diagnosis of gearboxes

Unsupervised domain adaptation alleviates the dependencies of conventional fault diagnosis methods on sufficient labeled data and strict data distributions. Nonetheless, the current domain adaptation methods only concentrate on the data distributions and ignore the feature gradient distributions, leading to some samples being misclassified due to large gradient discrepancies, thus affecting diagnosis performance. In this paper, a gradient aligned domain adversarial network (GADAN) is proposed. First, the discrepancies of the marginal and conditional distribution between the source and target domain are reduced by minimizing the joint maximum mean discrepancy. Then, a pseudo-labeling approach based on a clustering self-supervised strategy is utilized to attain high-quality pseudo-labels of target domains, and most importantly in the dimension of the data gradient, the feature gradient distributions are aligned by adversarial learning to further reduce the domain shift, even if the distributions of the two domains are close enough. Finally, experiments and engineering applications demonstrate the effectiveness and superiority of GADAN for transfer diagnosis between various working conditions or different machines.

Gearbox Fault Diagnosis Based on ICEEMDAN-MPE-AWT and SE-ResNeXt50 Transfer Learning Model

Because the gearbox in transmission systems is prone to failure and the fault signal is not obvious, the fault end cannot be located. In this paper, a gearbox fault diagnosis method grounded on improved complete ensemble empirical mode decomposition with adaptive noise, a multiscale permutation entropy and adaptive wavelet thresholding (ICEEMDAN-MPE-AWT) denoising method and an SE-ResNeXt50 transfer learning model are proposed. Initially, the vibration signal is denoised by ICEEMDAN-MPE-AWT, the denoised vibration signal is then converted into a Gram angle field (GAF) diagram, and then the parameters are transferred by the fine-tuning transfer learning strategy. Finally, a GAF diagram is input into the model for training to achieve fault extraction and classification. In this paper, the open gear dataset of Southeast University is used for experimental research. The experimental results show that when using the ICEEMDAN-MPE-AWT and when the signal-to-noise ratio (SNR) of the experimental data is −4 dB, the average accuracy of the GASF+TSE-ResNeXt50 and the GASF+TSE-ResNeXt18 can reach 98.8% and 97.5%, respectively. When the SNR is 6 dB, the accuracy of the above two models reaches 100% and 99.3%, respectively. Moreover, when compared to alternative approaches, the noise reduction method in this paper can better remove noise interference so that the model can better extract fault features. Therefore, the method proposed in this article shows significant improvement in noise reduction and fault classification accuracy compared to other methods.

A multi-scale graph convolutional network with contrastive-learning enhanced self-attention pooling for intelligent fault diagnosis of gearbox

Recently, the emerging graph convolutional network (GCN) has been applied into fault diagnosis with the aim of providing additional fault features through topological information. However, there are some limitations with these methods. First, the interactions between multi-frequency scales are ignored in existing studies, while they mainly focus on constructing graphs through the relationship between channels/instances. Second, the constructed graph cannot well reflect the topology of noisy samples and lacks robust hierarchical representation learning capability, and the learned graphs have limited interpretability. Hence, a Multi-Scale GCN with Contrastive-learning enhanced Self-attention Pooling (MSGCN-CSP) method is proposed for intelligent fault diagnosis of gearbox. Time–frequency distributions are converted into multi-scale graphs to extract fault features through topological relationships between multi-frequencies. Contrastive-learning is used to implement graph pooling, which enables hierarchical representation learning. Experimental results on two gearbox datasets illustrate that the proposed method offers competitive diagnostic performance and provides good interpretability in establishing GCN.

A TLOT train gearbox fault diagnosis method based on ridge extraction under variable speed conditions

Owing to the rapidly varying working conditions of urban rail trains, the rotational speed conditions constantly shift in a short time span. As a key component of the running gear, the gearbox generates non-stationary vibration signals, making it challenging to monitor its health status. To address this challenge, a tacholess order tracking method (TLOT) based on ridge extraction method is proposed in this paper. This method determines the optimal search starting point using the time-frequency Gini coefficient and extracts the time-varying gearbox meshing frequency components from the time-frequency representation results. Furthermore, the TLOT method is utilized to process the strongly time-varying gearbox signal. In the order domain, multiple frequency components are extracted to reconstruct the signal. Simulation and experimental results demonstrate that the proposed method achieves a superior ridge extraction effect on gearbox experimental signals. It accurately converts unstable signals into angular domain stable signals, enhancing the energy aggregation of time-varying unstable signals in the order domain. This approach addresses the problem of weak harmonic component extraction from gearbox signals and effectively reduces interference in the resonance band, realizing fault diagnosis of the gearbox under unstable working conditions.

Multi-modal data cross-domain fusion network for gearbox fault diagnosis under variable operating conditions

Gearbox fault diagnosis is a critical aspect of machinery maintenance and reliability, as it ensures the safe and efficient operation of various industrial systems. The cross-domain fault diagnosis method based on transfer learning has been extensively researched to enhance the engineering applications of data-driven methods. Currently, the state-of-the-art gearbox cross-domain fault diagnosis primarily relies on single-modal data, which may not capture the full information needed for robust fault diagnosis under varying conditions. To address this issue, we propose a novel multi-modal data cross-domain fusion network that utilizes vibration signals and thermal images to capture comprehensive information about the gearbox’s health conditions. First, one-dimensional and two-dimensional convolutional neural networks are constructed for feature extraction and fusion of multi-modal data. Then, the Maximum Mean Discrepancy loss is introduced to achieve cross-domain feature alignments within the modal. Finally, the cross-modal consistency learning strategy is constructed to enhance the cross-domain diagnosis performance of the model. To validate the effectiveness of the proposed method, we conducted experiments on a real-world gearbox test rig. Experimental results demonstrate that the proposed method is superior to single-modal methods and existing fusion methods in terms of diagnosis performance, proving that the proposed method offers a promising solution for gearbox fault diagnosis in industrial settings.

Deep continuous convolutional networks for fault diagnosis

Convolutional neural network (CNN) architectures have been extensively utilized in data-driven fault diagnosis and have demonstrated significant success. However, there remain certain constraints or restrictions of standard CNN for real-world applications, including discrete convolutions with a priori kernel size that failed to capture richer long-term features, performance deterioration when testing on irregular-length signals, and low accuracy under imbalanced datasets with severe levels of noise. In this paper, a new deep continuous convolutional network (DCCN) based on implicit neural representations is proposed to alleviate these limitations. Multilayer perceptrons with Gaussian masks are developed for the parameterization of continuous convolutional kernels, while multiplicative Gabor filters are used to improve the anti-noise ability. Depth-wise separable convolutions with residual connections are used to improve the computational efficiency of DCCN. The feasibility of DCCN has been verified on two laboratory datasets and one public dataset. The proposed approach achieves average macro F1-scores of over 98.69 % for bearing, gearbox, and mixed fault diagnosis under different noise interferences. Moreover, DCCN achieves an average macro F1-score of 86.43 % when training on 2048 data points and testing on 512 data points. Results show that DCCN has more robust recognition and anti-noise capability than state-of-art methods.

Fault diagnosis of gearbox driven by vibration response mechanism and enhanced unsupervised domain adaptation

Although data-driven model has achieved remarkable results in gearbox fault diagnosis, its diagnostic accuracy is still highly dependent on large amounts of high-quality labeled samples. Some data generation methods, such as generative adversarial network, are utilized to address this problem. However, the generated simulation samples not only lack fault mechanism features with clear physical meaning, but also have distribution differences with the real samples. Aiming at the above problems, an enhanced unsupervised domain adaption method combined with vibration response mechanism is proposed for gearbox fault diagnosis. Firstly, various fault types of labeled simulation signals with clear physical meaning are generated based on vibration response mechanism of gearbox, alleviating the lack of large amounts of high-quality labeled samples for data-driven models. Secondly, to narrow the inevitable domain discrepancy between simulation samples and experimental samples, a domain mapping method is raised to both transform their distributions to normal distribution by optimizing an alignment function, which also could effectively improve the diagnostic speed and accuracy of intelligent models. Finally, the mapped samples are directly fed into an arbitrary unsupervised domain adaptation model to achieve fault diagnosis in the absence of any label information of measured samples. Importantly, the proposed domain mapping method can be simply appended to any existing core network to enhance diagnostic accuracy without necessitating modifications to its structure or training procedure. Experiments on two gearbox datasets suggest that the proposed method can effectively boost the performance of diagnosis issues with only a small number of experimental samples and outperform existing diagnosis approaches.

Gearbox fault diagnosis method based on improved semi-supervised MTDL and GAF

Aiming at the problem that it takes a long time and high cost to obtain complete labeled data under intelligent fault diagnosis and unlabeled data is not used. This paper proposes an improved semi-supervised mean teacher deep learning (MTDL) and Gramian angle field (GAF) fusion diagnostic method. This method fully utilizes a small number of labeled samples and a large number of unlabeled samples to deeply mine invisible fault features and potential physical correlations. At the same time, it solves the problem of losing the inter-data correlation structure when one-dimensional time series signals are used as inputs for neural networks. The GAF-MTDL method uses consistency regularization and modifies the network structure in the mean teacher algorithm into a semi-supervised deep learning model enhanced by WideResNet. The experimental results show that the proposed GAF-MTDL method saves a lot of manual labeling costs, improves the recognition accuracy and generalization ability, and can achieve excellent prediction accuracy with very little labeled data. In the end, the accuracy of planetary gear fault identification reached 98.22% under the labeling rate of 20%, and the accuracy of fault identification reached 99.98% through the verification of the bearing data set of Case Western Reserve University. The value of this research is to bring an efficient and low-cost technology to the field of industrial intelligent fault diagnosis, which can significantly improve the accuracy of fault identification.

Human-computer Interaction Perspectives on Condition Monitoring and Fault Diagnosis of Gearbox in Doubly-Fed Wind Turbines: An Exploration with Artificial Intelligence Technology

Human-computer Interaction Perspectives on Condition Monitoring and Fault Diagnosis of Gearbox in Doubly-Fed Wind Turbines: An Exploration with Artificial Intelligence Technology