Difficulty Of Fault Diagnosis Research Articles

Fault diagnosis of rolling bearings based on dynamic convolution and dual-channel feature fusion under variable working conditions

Addressing the challenge of inconsistent data feature distribution and the difficulty of fault diagnosis in rolling bearings operating under variable conditions, a novel approach is proposed for bearings fault diagnosis. Dynamic convolution and dual-channel feature fusion are utilized in our method. In the shallow network layer, we employ a dual-channel convolutional structure, combining a large convolutional group with a small convolutional group to enhance the extraction of high-low frequency fault information from images. The improved GhostNetV2 bottleneck layer was used in the deeper layer of the network to obtain more beneficial features through the dynamic convolution and attention mechanism. Finally, fault classification and evaluation under variable working conditions was performed on the Case Western Reserve University and Drivetrain Dynamic Simulator (DDS) datasets. Our results showed that the methods and model used in this study can effectively handle the precision fault detection across various operational scenarios.

Automatic fault diagnosis of rolling bearings under multiple working conditions based on unsupervised stack denoising autoencoder

In practical engineering, data often lack labels, resulting in difficulty in fault diagnosis. Because stack-denoising autoencoders possess robust feature extraction capabilities and resistance to interference, an automatic and unsupervised bearing fault diagnosis method based on the stack-denoising autoencoder without an output layer was proposed in this study. As the stacked denoising autoencoder is an unsupervised algorithm, this approach can reduce reliance on manually labeled data labels. Therefore, this study proposed a new method for automatic fault diagnosis. First, the bearing fault features of the rolling bearing were extracted using the stack denoising autoencoder without an output layer. Meanwhile, the dimensions of the features were directly reduced to two or three dimensions by several hidden layers, thereby reducing manual experience. Second, the labels extracted from the clustering model were selected as inputs for different classifier models to automatically identify different types of faults. Two open-source rolling bearing datasets under various conditions were used to validate the classification performance of the proposed method. Finally, its effectiveness was verified using the experimental results. Various indicators were used to evaluate the performance of the proposed method, and the results showed an automatic bearing fault diagnosis accuracy of up to 90% when using different models and working conditions. Among the two datasets, the classification model achieved the highest accuracies of 0.99667 and 0.97143 and the lowest accuracies of 0.98000 and 0.90476, respectively.

A Diagnosability-Integrated Design Approach Based on Graph Theory

In order to take into account the influence of both system structure and diagnosis algorithm in the diagnosability design of the system, a diagnosability-integrated design method based on graph theory was proposed in this paper. Firstly, based on the diagnosability evaluation results, the difficulty of fault diagnosis was qualitatively analyzed using the K-means method, and the diagnosis plot of measurement point was drawn based on the analysis results. Secondly, the Bron–Kerbosch algorithm was used to extract the maximal cliques from the diagnosis plot of measurement point and determine the set of maximal cliques that can diagnose faults in the system based on the hypergraph edge coverage theorem. Finally, a cascade classifier was set on the maximal clique set to classify and identify faults in the system, and the performance of the diagnosis scheme was evaluated using the posterior probabilities of the classifier outputs combined with the Shannon entropy. At the same time, the method incorporated a measurement point update mechanism, which can decide whether to add additional measurement point according to the evaluation results of Shannon entropy to ensure better diagnosis effect. The results of simulation experiments showed that the fault diagnosis scheme designed by the method of this paper improved the correct rate of diagnosis results by 3.25 percentage points compared with other diagnosis schemes due to the simultaneous consideration of the structure of the system and the diagnosis method, and the diagnosis results of this paper were relatively stable in repeated experiments, which proved the practicality and effectiveness of the method of this paper.

Fault Diagnosis of Wind Turbine Generators Based on Stacking Integration Algorithm and Adaptive Threshold.

Fault alarm time lag is one of the difficulties in fault diagnosis of wind turbine generators (WTGs), and the existing methods are insufficient to achieve accurate and rapid fault diagnosis of WTGs, and the operation and maintenance costs of WTGs are too high. To invent a new method for fast and accurate fault diagnosis of WTGs, this study constructs a stacking integration model based on the machine learning algorithms light gradient boosting machine (LightGBM), extreme gradient boosting (XGBoost), and stochastic gradient descent regressor (SGDRegressor) using publicly available datasets from Energias De Portugal (EDP). This model is automatically tuned for hyperparameters during training using Bayesian tuning, and the coefficient of determination (R2) and root mean square error (RMSE) were used to evaluate the model to determine its applicability and accuracy. The fitted residuals of the test set were calculated, the Pauta criterion (3σ) and the temporal sliding window were applied, and a final adaptive threshold method for accurate fault diagnosis and alarming was created. The model validation results show that the adaptive threshold method proposed in this study is better than the fixed threshold for diagnosis, and the alarm times for the GENERATOR fault type, GENERATOR_BEARING fault type, and TRANSFORMER fault type are 1.5 h, 5.8 h, and 3 h earlier, respectively.

Compound Fault Diagnosis for Rotating Machinery: State-of-the-Art, Challenges, and Opportunities

Compound fault, as a primary failure leading to unexpected downtime of rotating machinery, dramatically increases the difficulty in fault diagnosis. To deal with the difficulty encountered in implementing compound fault diagnosis (CFD), researchers and engineers from industry and academia have made numerous significant breakthroughs in recent years. Admittedly, many systematic surveys focused on fault diagnosis have been conducted by reputable researchers. Nevertheless, previous review articles paid more attention to fault diagnosis with several single or independent faults, resulting in that there is still lacking a comprehensive survey on CFD. Therefore, to fulfill the above requirements, it is necessary to provide an in-depth overview of fault diagnosis methods or algorithms for compound faults of rotating machinery and uncover potential challenges or opportunities that would guide and inspire readers to devote their efforts to promoting fault diagnosis technology more effective and practical. Specifically, the backgrounds including the related definitions and a new taxonomy of CFD methods are detailed according to the way of implementing compound fault recognition. Then, the state-of-the-art applications of CFD are overviewed based on relevant publications in the past decades. Finally, the challenges and opportunities associated with implementing CFD are concluded and followed by a conclusion for ending this survey. We believe that this review article can provide a systematic guideline of CFD from different aspects for potential readers and seasoned researchers.

A Fault Diagnosis Method for PV Arrays Based on New Feature Extraction and Improved the Fuzzy C-Mean Clustering

Photovoltaic (PV) array fault diagnosis is vital for the safe and stable operation of PV systems. Up to now, there are many methods to diagnose and classify PV array faults successfully. However, the difficulty of fault diagnosis is increased because the PV arrays have nonlinear output characteristics and complex working environments. In practice, the difference in characteristic parameters of different faults is not apparent, and it is not easy to effectively obtain labels of many samples. In order to address the above problems, this article proposes a new fault detection method for PV arrays based on the output current–voltage ( <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">I</i> – <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">V</i> ) characteristic, an improved fuzzy C-mean clustering (FCM) algorithm to identify four common PV array faults. The measured <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">I</i> – <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">V</i> characteristic curves are used to extract the initial feature parameters and then calculate the initial parameters to obtain characteristic parameters. In addition, it used characteristic parameters as feature variables of the FCM fault diagnosis model. Finally, classification results are verified by inner cluster maximum mean discrepancy and reclassification pseudoparameter based on the relationship between characteristic parameters. This method defines new characteristic parameters and achieves the purpose of fault detection by reclassification; in addition, we verify the high accuracy and simplicity of the method through simulation and experiment.

A Novel Wind Turbine Health Condition Monitoring Method Based on Correlative Features Domain Adaptation

Aimed at the difficulty in fault diagnosis of wind turbine transmission system under variable working conditions, the paper proposes a novel health condition monitoring method based on correlative features domain adaptation. Firstly, the envelope analysis of the collected signals is carried out, and the time–frequency features of the signals are extracted to construct the feature set. The feature sets under the similar working conditions to target are selected as the auxiliary sample sets in source domain through the transferability evaluation. Then, a transformation matrix is found to adapt the marginal and conditional distributions of wind turbine sample data under different working conditions, and its weight is adjusted. While reducing the discrepancy between domains, the class imbalance problem is taken into consideration, so as to improve the accuracy of fault diagnosis under the target working condition. Finally, the classifier is trained with the adjusted source domain and tested in the target domain. Experiments show that the proposed method can effectively improve the accuracy of wind turbine fault diagnosis.

Dissipativity-based distributed fault diagnosis for plantwide chemical processes

Most modern chemical processes consist of a number of process units interconnected with mass and energy flows, often with energy integration and materials recycle loops. As such, faults (process faults, actuator faults, or sensor faults) often propagate to multiple process units (subsystems), causing significant difficulties in fault diagnosis for plantwide systems. In this paper, a general distributed fault diagnosis approach is proposed for plantwide chemical processes, which takes into account the interactions among process units. The distributed fault diagnostic observers are designed to be sensitive to the local faults (local sensitivity) and insensitive to faults in other process units (remote faults insensitivity) and disturbances. The above requirements are formulated as plantwide dissipativity conditions and the gains for the distributed estimators and residual generators are obtained offline by solving a set of linear matrix inequalities. A case study of heat exchanger network is presented to demonstrate the effectiveness of the proposed approach.

Novel Multistate Fault Diagnosis and Location Method for Key Components of High-Speed Trains

Images collected by linear scan cameras are stretched and compressed due to the speed regulation of trains. This condition changes the shape of objects and considerably increases the false and missed alarm rates. Moreover, small size and dense distribution of the key components in railway trains increase the difficulty in fault diagnosis. Therefore, a two-module troubleshooting and positioning methodology is proposed, in this article, for state diagnosis of key small components of high-speed running trains. First, the image with deformation is reshaped to be exactly the same as the standard one via omnidirectional scale correlation normalization (OSCN), which performs well even in low texture, high-light situations. Second, we propose three feature-enhanced models to expand the receptive field of deep feature maps. This novel detector, namely, refine-inception net (RIN), not only reduces the rate of missed and false detection, but also minimizes the influence of target size and occlusion. Augmentation is performed to increase robustness because the detector is data driven. Experimental results show that the optimal strategy combining OSCN and RIN can troubleshoot high-speed trains of different models with an accuracy higher than 99%. Our method can be extended to foreign object recognition on the train roof while maintaining railway safety.

Intelligent Fault-Diagnosis System for Acoustic Logging Tool Based on Multi-Technology Fusion.

To improve the performance of acoustic logging tool in detecting three-dimensional formation, larger and more complicated transducer arrays have been used, which will greatly increase the difficulty of fault diagnosis during tool assembly and maintenance. As a result, traditional passive diagnostic methods become inefficient, and very skilled assemblers and maintainers are required. In this study, fault-diagnosis requirement for the acoustic logging tool at different levels has been analyzed from the perspective of the tool designer. An intelligent fault-diagnosis system consisting of a master-slave hardware architecture and a systemic diagnosis strategy was developed. The hardware system is based on the embedded technology, while the diagnosis strategy is built upon fault-tree analysis and data-driven methods. Diagnostic practice shows that this intelligent system can achieve four levels of fault diagnosis for the acoustic logging tool: System, subsystem, circuit board, and component. This study provided a more rigorous and professional fault diagnosis during tool assembly and maintenance. It is expected that this proposed method would be of great help in achieving cost reduction and improving work efficiency.

Research on Weak Fault Extraction Method for Alleviating the Mode Mixing of LMD.

Compared with the strong background noise, the energy entropy of early fault signals of bearings are weak under actual working conditions. Therefore, extracting the bearings’ early fault features has always been a major difficulty in fault diagnosis of rotating machinery. Based on the above problems, the masking method is introduced into the Local Mean Decomposition (LMD) decomposition process, and a weak fault extraction method based on LMD and mask signal (MS) is proposed. Due to the mode mixing of the product function (PF) components decomposed by LMD in the noisy background, it is difficult to distinguish the authenticity of the fault frequency. Therefore, the MS method is introduced to deal with the PF components that are decomposed by the LMD and have strong correlation with the original signal, so as to suppress the modal aliasing phenomenon and extract the fault frequencies. In this paper, the actual fault signal of the rolling bearing is analyzed. By combining the MS method with the LMD method, the fault signal mixed with the noise is processed. The kurtosis value at the fault frequency is increased by eight-fold, and the signal-to-noise ratio (SNR) is increased by 19.1%. The fault signal is successfully extracted by the proposed composite method.

Multi-label spacecraft electrical signal classification method based on DBN and random forest.

In spacecraft electrical signal characteristic data, there exists a large amount of data with high-dimensional features, a high computational complexity degree, and a low rate of identification problems, which causes great difficulty in fault diagnosis of spacecraft electronic load systems. This paper proposes a feature extraction method that is based on deep belief networks (DBN) and a classification method that is based on the random forest (RF) algorithm; The proposed algorithm mainly employs a multi-layer neural network to reduce the dimension of the original data, and then, classification is applied. Firstly, we use the method of wavelet denoising, which was used to pre-process the data. Secondly, the deep belief network is used to reduce the feature dimension and improve the rate of classification for the electrical characteristics data. Finally, we used the random forest algorithm to classify the data and comparing it with other algorithms. The experimental results show that compared with other algorithms, the proposed method shows excellent performance in terms of accuracy, computational efficiency, and stability in addressing spacecraft electrical signal data.

Fault Diagnosis of Roadheader Based on Dynamic Fault Tree

The roadheader faults are highly concerned by machinery factories. Facing the variety of the roadheader faults, the complex relationships between fault sources, as well as the difficulties of fault diagnosis, this paper proposed a method to set up the dynamic fault tree of roadheader. Modular method is used to split the fault tree into dynamic module and static module. Specifically, the binary decision diagram (BDD) method is used in static analyze module, while the dynamic module is built up considering the logical relationship between different faults thoroughly. This method combines the benefits of both static module and dynamic module to figure out the formation of faults. It is valuable in engineering to give guidance in the reliability design of roadheaders.

Frequency response model and mechanism for wind turbine planetary gear train vibration analysis

Frequency components of complicated asymmetric modulation sidebands, existing in the vibration of the healthy planetary gear train, are prone to be erratically diagnosed as fault characteristics, which leads to difficulties in fault diagnosis of wind turbine planetary gearbox. The factors affecting the modulation sideband, i.e. the periodical time-varying transmission path and meshing force direction, are analysed. Considering both the meshing vibrations of the planet–ring and planet–sun gear pairs, a mathematical model was developed to analyse the planetary gear train's vibration response. Simulation and experiments were conducted, and the mechanism of vibration modulation sidebands was revealed. The modulation sideband is not caused by the meshing vibration itself, but by the testing method that sensors are fixed on the ring gear or gearbox casing. The frequency components and amplitudes of the sidebands are determined by the tooth number of the ring gear and sun gear, the number of planet gears and their initial assembling phases. The asymmetric modulation sideband is mainly caused by the phase difference of the initial planets' assembling phase.

Fault Modeling and Testing for Analog Circuits in Complex Space Based on Supply Current and Output Voltage

This paper deals with the modeling of fault for analog circuits. A two-dimensional (2D) fault model is first proposed based on collaborative analysis of supply current and output voltage. This model is a family of circle loci on the complex plane, and it simplifies greatly the algorithms for test point selection and potential fault simulations, which are primary difficulties in fault diagnosis of analog circuits. Furthermore, in order to reduce the difficulty of fault location, an improved fault model in three-dimensional (3D) complex space is proposed, which achieves a far better fault detection ratio (FDR) against measurement error and parametric tolerance. To address the problem of fault masking in both 2D and 3D fault models, this paper proposes an effective design for testability (DFT) method. By adding redundant bypassing-components in the circuit under test (CUT), this method achieves excellent fault isolation ratio (FIR) in ambiguity group isolation. The efficacy of the proposed model and testing method is validated through experimental results provided in this paper.

Method Research on Information Fusion of Reciprocating Machinery Fault Diagnose

The reciprocating machinery has complicated structure, various motion forms and high difficulty of fault diagnosis, now we mainly apply the multi-sensor information fusion as the effective solution. This thesis introduces the structure model of reciprocating machinery information fusion, establishes the tree diagnosis model for fault diagnosis feature information fusion of reciprocating engine, describes the fusion diagnosis steps, and puts forward the technological approaches and methods to solve fault diagnosis information fusion of reciprocating engine from the aspect of testing methods and diagnosis modeling.

Joint amplitude and frequency demodulation analysis based on local mean decomposition for fault diagnosis of planetary gearboxes

The vibration signals of faulty planetary gearboxes have complicated spectral structures due to the amplitude modulation and frequency modulation (AMFM) nature of gear damage induced vibration and the additional multiplicative amplitude modulation (AM) effect caused by the time-varying vibration transfer paths (for local gear damage case) and the passing planets (for distributed gear damage case). The spectral complexity leads to the difficulty in fault diagnosis of planetary gearboxes. Observing that both the amplitude envelope and the instantaneous frequency of planetary gearbox vibration signals are associated with the characteristic frequency of the faulty gear, a joint amplitude and frequency demodulation method is proposed for fault diagnosis of planetary gearboxes. In order to satisfy the mono-component requirement by instantaneous frequency estimation, a signal is firstly decomposed into product functions (PF) using the local mean decomposition (LMD) method. Then, the earliest extracted PF that has an instantaneous frequency fluctuating around the gear meshing frequency or its harmonics is chosen for further analysis, because it contains most of the information about the gear fault. The amplitude demodulation analysis can be accomplished through Fourier transforming the amplitude envelope of the chosen PF. For the frequency demodulation analysis, Fourier transform is applied to the estimated instantaneous frequency of the chosen PF to reveal its fluctuating frequency, thus obtaining the spectrum of the instantaneous frequency. By joint application of the amplitude and frequency demodulation methods, planetary gearbox faults can be diagnosed by matching the dominant peaks in the envelope spectrum and the spectrum of instantaneous frequency with the theoretical characteristic frequencies of faulty gears. The performance of the proposed method is illustrated by simulated signal analysis, and is validated by experimental signal analysis of a lab planetary gearbox with intentionally created pitting and naturally developed wear.

Fault Diagnosis of Tolerance Analog Circuit Based on Wavelet Neural Network with PSO Algorithm

For the Difficulties in fault diagnosis of tolerance analog circuit, a Wavelet Neural Network (WNN) diagnosis method based on Particle Swarm Optimization (PSO) algorithm is proposed. To overcome the deficiencies of the traditional BP algorithm using in WNN, PSO algorithm is introduced into the parameters optimization in WNN, and the velocity disturbance operator is embedded to ensure the particle out of the premature position for PSO algorithm performance. The simulation results show that the proposed method has the fast training rate, accurate diagnosis, without local convergence.

Fault Diagnosis of Bearing in Mechanical Drive System Using Wavelet Analysis

Exploring a proper wavelet base function to characterize fault information of mechanical equipment is a big problem when wavelet analysis is applied into the field of fault diagnosis.Improper wavelet base function can dilute fault information, which will cause difficulties in fault diagnosis , based on the principle that the wavelet domain oscillation waveform are similar to the detected signal components, a proper wavelet base function was selected; through wavelet transform on the acceleration signals of the fault bearing, the spectrum chart of fault signals and the frequency characteristics of fault bearing were obtained. The experimental result shows that this method has a great advantage over extracting character signals.

Researches on New Fault Diagnosis Method for Process Industry Based on Improved SVM Reduction Algorithm

Once faults happened in the process industry, it will bring on heavy casualties and fatal property losses, even lead to irreclaimable pollution. Combining the large-scale sample set reduction strategy with SVM, this paper proposed a new improved support vector machine (SVM) training algorithm based on support vectors to solve the difficulty of fault diagnosis of process industry. A new reduction algorithm is presented to reduce large-scale sample set. The constraint threshold is introduced in the new algorithm. The new training algorithm can eliminate the samples correspond to non-support vectors so that poor generation ability caused by large-scale training sample set is solved. A new evaluating indicator of large-scale sample set reduction strategy is also proposed in this paper. The improved SVM algorithm proposed is applied to two typical fault diagnosis of process industry. The results of experiments show that the improved SVM algorithm not only reduces greatly the cost of SVM learning but also markedly increases the speed of classification, and at one time the accuracy of fault diagnosis is not debased.