Vibration Signals For Fault Diagnosis Research Articles

Transferable mapping shift network for unsupervised domain adaptation using in vibration signal fault diagnosis under variable conditions

Transferable mapping shift network for unsupervised domain adaptation using in vibration signal fault diagnosis under variable conditions

Fine-Gained Recurrence Graph: Graphical Modeling of Vibration Signal for Fault Diagnosis of Wind Turbine

Benefiting from the recent successes of convolutional neural networks (CNNs), many studies have modeled the vibration signal of energy system into a two-dimensional (2D) input graph to amplify and highlight fault features. However, most works seldomly consider the system dynamic characteristics, which may affect the knowledge discovery and diagnosis quality. To address this issue, this paper proposes a novel graphical modeling approach, termed as fine-gained recurrence graph (FRG), to capture dynamic characteristics of vibration signals from the view of nonlinear dynamics. FRG focuses on modeling the temporal correlations and tendencies between state vectors in phase space and then visualizes the characteristics to represent intrinsic dynamic changes of the vibration signals into 2D graphs. The information representation capability of the proposed FRG is verified using different dynamic signals. Based on this finding, an ensemble fault diagnosis model is proposed fusing FRG with deep CNNs. Meanwhile, transfer learning is accompanied to deal with the training difficulties of deep CNNs. Finally, case studies on experimental data and real wind turbine data illustrate its effectiveness and feasibility. Comparisons with four state-of-the-art approaches have confirmed the preferable information representation and diagnosis performance of the proposed approach.

Optimization Application of Deep Belief Network in Gear Fault Diagnosis

Deep belief networks (DBNs) in the process of the application of the gear vibration signal fault diagnosis, vector network depth, the number of neurons in hidden layer, set the initial parameters, such as there is a certain blindness, and for different data sets, the influence of initial parameters is different also, in order to get ideal fault recognition rate, the initial parameters have to be repeatedly debugging, This process is not only time-consuming and laborious, but also disadvantageous to its universal application in fault diagnosis. To solve this problem, this paper mainly optimizes the number of hidden layer neurons and learning rate of DBNs. The diagnosis example shows that after optimization, the initial parameter setting is more accurate and efficient, and the fault identification rate is significantly improved.

Fault feature extraction and recognition of vibration signal based on CWD-SK and HMM

Aiming at the problem of rolling bearing fault diagnosis caused by interference information in the process of vibration transmission, this paper puts forward the theory and method of vibration signal fault diagnosis based on Choi Williams distribution spectral kurtosis (SK) combined with hidden Markov model (HMM). The fault location method of local mean decomposition (LMD) is used to screen the PF component with obvious fault characteristic information, CWD-SK is calculated and set as the characteristic parameter to realize the feature extraction of the initial fault of the vibration signal. CWD-SK is input into HMM as the feature vector to classify and identify the fault features of the vibration signal. Finally, the effectiveness of the algorithm is verified by the simulation of the actual data. It provides a robust theoretical and methodological basis for the establishment of fault diagnosis and classification recognition technology suitable for initial rolling mechanical vibration signals.

Research on a Denoising Method of Vibration Signals Based on IMRSVD and Effective Component Selection

This paper proposes a denoising method of vibration signal based on improved multiresolution singular value decomposition (IMRSVD) and effective component selection. A new construction method of trajectory matrix is used, which can enhance the oscillating component of the original signal. Next, based on the improved trajectory matrix, singular value decomposition (SVD), which plays the role of pre-decomposition, is used to obtain multiple one-dimensional components, and the further decomposition of that is achieved by multiresolution singular value decomposition (MRSVD). Finally, the effective components selection of a series of decomposed signal components is achieved based on the proposed feature evaluation index (FEI). The denoising experiments are carried out using the simulation signal and the vibration signal of planetary gear, respectively. The experimental results show that the proposed method performs better than the traditional SVD denoising method, and the weak fault feature in the vibration signal can be extracted successfully. In addition, the comparison between periodic modulation intensity (PMI) and FEI displays that the proposed method has better robustness and accuracy than the interference components with similar frequency. Thus, the proposed method is an effective weak fault feature extraction and denoising tool of vibration signals for fault diagnosis.

A planetary gearbox fault diagnosis method based on time-series imaging feature fusion and a transformer model

As a crucial component in the transmission system, a planetary gearbox has a relatively complicated structure and usually operates under complex working conditions and a severe noisy environment, making it challenging to achieve precise and efficient fault diagnosis. Along with the development of artificial intelligence techniques, end-to-end fault diagnosis frameworks have been widely studied, among which convolutional and recurrent neural networks are the mainstream backbone networks. However, these networks have shortcomings in computational efficiency and feature extraction, which lead to the application of a self-attention mechanism. This paper presents a fault diagnosis method based on frequency domain Gramian angular field (GAF) and Markov transition field (MTF) features for planetary gearboxes by combining the characteristics of vibration signal fault diagnosis and transformer network structure. The experiments show that the frequency domain GAF-MTF features can effectively reduce the influence of time shifting between samples and improve diagnostic accuracy. Furthermore, comparisons with other mainstream models indicate that the proposed method can obtain competitive results and achieve more accurate and robust performance under noisy conditions.

Bearing Vibration Signal Fault Diagnosis Based on LSTM-Cascade CatBoost

<p>Bearing is one of the most concerned parts in the field of fault diagnosis. At present, there are numerous excellent algorithms applied to bearing fault detection. This paper proposes a new fault bearings diagnosis model named LSTM-Cascade CatBoost, which can directly classify bearing vibration signals in the case of multiple granularity and high dimensions without signal processing. The model is based on gcForest, whose complexity can be adjusted automatically to the size of data set and it uses LSTM to extract features of time series signals instead of multi-grained scanning for improving the model’s feature extraction ability. CatBoost is used as the base classifier of cascade forest to improve the classification accuracy of the model. Experimental results show the fact that this model is highly accurate in CWRU and XJTU-SY datasets. Besides, it not only proves that the feature extraction ability of LSTM is significantly better than that of multi-grained scanning, but CatBoost as a base classifier can further improve the accuracy of cascade forest.</p> <p> </p>

End-to-End Continuous/Discontinuous Feature Fusion Method with Attention for Rolling Bearing Fault Diagnosis

Mechanical equipment failure may cause massive economic and even life loss. Therefore, the diagnosis of the failures of machine parts in time is crucial. The rolling bearings are one of the most valuable parts, which have attracted the focus of fault diagnosis. Many successful rolling bearing fault diagnoses have been made based on machine learning and deep learning. However, most diagnosis methods still rely on complex signal processing and artificial features, bringing many costs to the deployment and migration of diagnostic models. This paper proposes an end-to-end continuous/discontinuous feature fusion method for rolling bearing fault diagnosis (C/D-FUSA). This method comprises long short-term memory (LSTM), convolutional neural networks (CNN) and attention mechanism, which automatically extracts the continuous and discontinuous features from vibration signals for fault diagnosis. We also propose a contextual-dependent attention module for the LSTM layers. We compare the method with the other simpler deep learning methods and state-of-the-art methods in rolling bearing fault data sets with different sample rates. The results show that our method is more accurate than the other methods with real-time inference. It is also easy to be deployed and trained in a new environment.

A novel adaptive fault diagnosis algorithm for multi-machine equipment: application in bearing and diesel engine

This paper proposes an adaptive fault diagnosis algorithm based on vibration signals for fault diagnosis of bearings and diesel engines. First, the improved nonlinear gray wolf optimization algorithm (NGWO) is adopted to optimize the key parameter for variational mode decomposition (VMD) with the power spectral entropy as the fitness value. Meanwhile, adaptive noise reduction of the signal is realized. Then, sensitive fault features of bearings and diesel engines are selected through a feature sensitivity analysis on the vibration signals. Also, a single-layer sparse autoencoder is used to align the feature dimensions of each type of data to construct feature matrix samples. Subsequently, a deep neural network (DNN) consisting of a two-layer stacked sparse autoencoder (SSAE) and a Softmax classification layer is constructed to realize failure mode recognition. During the training process of DNN, a surrogate model formed by NGWO and a back propagation neural network is employed to optimize the hyperparameters of SSAE. Finally, to verify the effectiveness of the proposed fault diagnosis algorithm, fault diagnosis experiments are conducted on the fault data set of bearings and diesel engines. The diagnosis results show that the proposed method achieves high-precision fault diagnosis for bearings and diesel engines and performs stably for small samples.

A fusion CNN driven by images and vibration signals for fault diagnosis of gearbox

Abstract Gearbox diagnosis is critical for avoiding catastrophic failure and minimizing financial damages. Aiming at the problem that the vibration-based fault diagnosis methods cannot effectively identify the non-structural failure mode and the diagnosis model based on the infrared thermal image is not robust enough, a fusion fault diagnosis method for gearboxes using vibration signals and infrared images is proposed. By fusing these two kinds of heterogeneous data, the proposed method can identify both structural and unstructured health states while maintaining high robustness. In addition, CNN has powerful image processing capabilities, which can directly process two-dimensional infrared images and achieve high accuracy. Finally, a gearbox experiment is carried out to test the performance of our method. The results suggest that the proposed fusion CNN can obtain the highest accuracy compared with some methods based on single signals, shallow learning methods SVM and deep unsupervised learning methods SAE.

Fault diagnosis of rotating machinery based on combination of Wasserstein generative adversarial networks and long short term memory fully convolutional network

The traditional fault diagnosis methods of rotating machinery based on deep learning have made some achievements. However, the fault samples are generally difficult to collect, which will lead to data imbalance problem. In such cases, it will degrade their diagnosis performances. Therefore, a data rebalancing solution based on Wasserstein Generative Adversarial Network (WGAN) was proposed to improve data quality, and combined with long and short-term memory full convolutional network (LSTM-FCN) to achieve high-dimensional vibration signal fault diagnosis. The method first utilizes WGAN to expand the imbalanced fault samples. The unique one-dimensional convolution architecture in WGAN and the loss function based on Wasserstein distance ensure the stable feature learning ability of the model and the high quality of the generated data. Hence, it fully solves the problem of data imbalance. Subsequently, the balanced high-dimensional fault signals are fed into LSTM-FCN model. LSTM-FCN simultaneously combines the characteristics of LSTM and FCN in timing feature extraction, and thus extracts more robust and representative fault features through parallel feature extraction and fusion technology. Finally, SoftMax classifier is employed to identify different fault conditions. The proposed method is verified on the multiple unbalanced fault datasets of rolling bearings and gears. Compared with the traditional data imbalance resolution algorithms, the experimental results demonstrate that WGAN can effectively deal with the data imbalance problem, and the diagnosis accuracy of LSTM-FCN is maintained at about 99%. In comparison with the traditional fault diagnosis algorithms, the proposed method shows better diagnostic performance and generalization ability under complex working conditions with small samples and strong noise.

Fusion Domain-Adaptation CNN Driven by Images and Vibration Signals for Fault Diagnosis of Gearbox Cross-Working Conditions

The vibration signal of gearboxes contains abundant fault information, which can be used for condition monitoring. However, vibration signal is ineffective for some non-structural failures. In order to resolve this dilemma, infrared thermal images are introduced to combine with vibration signals via fusion domain-adaptation convolutional neural network (FDACNN), which can diagnose both structural and non-structural failures under various working conditions. First, the measured raw signals are converted into frequency and squared envelope spectrum to characterize the health states of the gearbox. Second, the sequences of the frequency and squared envelope spectrum are arranged into two-dimensional format, which are combined with infrared thermal images to form fusion data. Finally, the adversarial network is introduced to realize the state recognition of structural and non-structural faults in the unlabeled target domain. An experiment of gearbox test rigs was used for effectiveness validation by measuring both vibration and infrared thermal images. The results suggest that the proposed FDACNN method performs best in cross-domain fault diagnosis of gearboxes via multi-source heterogeneous data compared with the other four methods.

Feature extraction and classification algorithm, which one is more essential? An experimental study on a specific task of vibration signal diagnosis

With the development of machine learning, a data-driven model has been widely used in vibration signal fault diagnosis. Most data-driven machine learning algorithms are built based on well-designed features, but feature extraction is usually required to be completed in advance. In the deep learning era, feature extraction and classifier learning are conducted simultaneously, which will lead to an end-to-end learning system. This paper explores which one of the two key factors, i.e., feature extraction and classification algorithm, is more essential for a specific task of vibration signal diagnosis during a learning system is generated. Feature extractions from vibration signal based on both well-known Gaussian model and statistical characteristics are discussed, respectively. And several classification algorithms are selected to experimentally validate the comparative impact of both feature extraction and classification algorithm on prediction performance.

MLPC-CNN: A multi-sensor vibration signal fault diagnosis method under less computing resources

This paper proposes a fault diagnosis method for multi-sensor vibration signals under few computing resources, called multi-level feature fusion convolution neural network based on multi-layer pooling classifiers (MLPC-CNN). First, MLPC-CNN introduces the single-sensor-to-single-channel (STSSC) convolution to comprehensively extract features from multi-sensor data grayscale image that integrates all sensor information. This design can adopt more targeted filtering strategies for the samples from different sensors, and avoid the risk of extracting conflicting evidence. Second, MLPC-CNN uses a bypass branch structure based on average pooling layer. This design fuses different levels of signal features extracted by different layers without increasing the network learning parameters, which can extract high-level features while retaining more information from low-dimensional features. Third, MLPC-CNN introduces a multi-layer pooling classifier to replace the fully connected layer in traditional CNN. The pooling layers with different scales are used to achieve multiple functions, which greatly reduces the number of network parameters and the risk of overfitting. The measured data collected by the fault simulation test stand and the bearing fault dataset produced by case western reserve university are used to verify the performance of MLPC-CNN. Experimental results show that MLPC-CNN has reached 100% accuracy on both two datasets. In addition, to explore the fault diagnosis mechanism of MLPC-CNN, this paper uses multiple visualization methods to analyze the function of the convolution kernel in the STSSC convolution layer, the maximum activation feature signal of different convolution channels, and the evolution process of features generated from different fault samples.

Fault diagnosis of a planetary gearbox by D norm-based time synchronous averaging (DTSA) with roughly estimated phase information under an encoder-less operating condition

Minimum entropy deconvolution (MED) filter is widely used to extract fault-related impulse features from noisy vibration signals for fault diagnosis of a gearbox. However, the MED process is vulnerable to abrupt environmental and operational noise, as disturbance components can be extracted instead of fault-related features. To mitigate this issue, time-synchronous averaging (TSA) can be employed to reduce the remaining noise of the MED-filtered signal further. However, TSA requires accurately resampled vibration signals with the aid of an additional encoder system, which could be unavailable in the actual field. Although a typical phase estimation method can be employed, the phase estimation accuracy cannot be precise enough for TSA. In addition, TSA cannot fundamentally recover the fault-related features if the conventional MED filter is undesirably optimized to extract a single dominant impulse feature while eliminating fault-related features. To overcome these challenges, this paper proposes the novel use of D norms for TSA (DTSA). It was found in this paper that D norms, which encapsulate results from multiple MED filters, are amplified at every fault position with margins equal to the length of the MED filter, enabling Dnorm-based TSA (DTSA) by alleviating the slight phase error from vibration signals under an encoder-less condition. In addition, D norms contain all of the fault-related features from the measured signal and thus prevent undesirable eliminations of fault-related features as opposed to the conventional MED process. It was found in this paper that case studies involving an analytic simulation and a 2-kW planetary gearbox testbed indicate that the proposed method outperforms the conventional MED process under a slightly varying speed condition even without an encoder.

COMBINED SIGNAL PROCESSING METHOD FOR DIAGNOSIS AND MONITORING OF THE INDUCTION MOTORS OPTIMIZED FOR EMBEDDED SYSTEMS

Induction Motors (IM) play a key role in modern industry, so the condition monitoring systems are becoming increasingly relevant. Commercial monitoring systems are usually based on the measurement of IM’s vibrations and the further processing of the measured vibration signals. For those purposes the embedded systems (such as microcontrollers and inexpensive processors) are used. Embedded systems have limited resources, so data processing algorithms should have low computational complexity and require little memory. In this paper, the wellknown methods of processing vibration signals for fault diagnosis of the IM are considered and their main advantages and disadvantages for the implementation in embedded systems are highlighted. The previously proposed method based on a combination of the fast Fourier transform and the statistics of the fractional moments is optimized for vibration signal processing and implementation in embedded systems. The efficiency of diagnosis of such faults as eccentricity and a broke rotor bar, using the proposed method, is verified on the radial vertical vibrations measurements of the real motors under different constant load levels: no load, 50 % of the rated load, 75% of the rated load. The results show that this approach allows accurately diagnose the considered faults independently from the load level.

Research on distribution characteristics of vibration signals of ±500 kV HVDC converter transformer winding based on load test

The converter transformer is important for the high-voltage direct current transmission system. The distribution characteristics of the vibration signal of the tank are significant for the operation status monitoring and the judgment basis for the fault condition of the converter transformer. However, most present research on vibration detection technology focused on ordinary power transformers rather than the HVDC converter transformers. Moreover, the influence of the special structure and the heat accumulation effect of the load current on the vibration signal distribution characteristics is not fully considered. To solve the above problems, this paper analyzes the intensity distribution of vibration signals at different positions on the tank surface and valve side bushings based on the offline load test. At the same time, to study the influence factors of vibration signal distribution, the influence of current heat accumulation on winding characteristics and transformer oil temperature is analyzed. The results show that vibration intensity is related to location and the load current. And the vibration intensity of the tank surface increases with the time of current flowing through the winding, nevertheless, the vibration intensity tends to decrease near the valve-side bushing. The test methods and research on distribution Characteristics in this paper can provide new thoughts for the optimization of the layout of converter transformer vibration detection sensors, and provide a research basis for the application of vibration signal fault diagnosis methods.

Vibration fault diagnosis based on stochastic configuration neural networks

This work presents a study on fault diagnosis in vibration signal processing. Rather than building a fault model through frequently used approaches to handling the series data such as LSTM or hidden Markov field, this work processes the vibration signal by moving the time window to generate multiple samples and then transfers fault diagnosis into a traditional supervised learning problem. Stochastic configuration neural network (SCN) which gives a clear condition of guaranteeing high performance of randomly weighted neural networks is selected as the model for training and testing. Different classifiers are used to conduct a performance comparison, and their comparative advantages including why SCN particularly suitable for this type of learning and more discussions about the experimental results are shown. The paper provides a new scheme to processing vibration signal for fault diagnosis and some useful guidelines of building an appropriate model with high performance.

Finding Useful Features in Vibration Signals for Fault Diagnosis

AbstractThis contribution presents a selection of characteristic values derived from vibration signals that are useful for diagnosis of unbalance and coupling bend in hydropower rotors. The characteristic values are used with machine learning techniques for the estimation of fault type as well as fault intensity.

Fault diagnosis of vibration signal of synchronous condenser based on wavelet model

A large synchronous condenser has a complex structure and many parts, which is very prone to failure. However, the regular electrical quantity generally does not change significantly at the initial stage of failure, and may not reflect the fault characteristics. Some non-electrical quantities, such as vibration characteristics, can be used to diagnose and predict faults in a timely manner. Based on this, a multilayer-perceptron based wavelet model is proposed for fault diagnosis. Vibration signals are taken in real time at different positions on the surface of the condenser, and the energy eigenvalues are calculated by fast Fourier transform of the data, and the wavelet neural network model is input for training to obtain the nonlinear mapping relationship between the vibration signals and the fault types of the condenser, so as to realize the fault diagnosis of the condenser. Experimental results show that this method can effectively detect and diagnose the fault of the synchronous condenser in the early stage.