Noise Of Vibration Signal Research Articles

Noise Reduction Method for the Vibration Signal of Reactor CRDM Based on CEEMDAACN-SK

The reactor control rod drive mechanism (CRDM) controls the startup, shutdown and power of the reactor; it is one of the key pieces of equipment to ensure the normal operation of the reactor. CRDM is complex, mainly composed of stator, rotor, bearing, roller, etc. The characteristic analysis of the vibration monitoring signal is one of the important methods of the CRDM state evaluation. In view of the characteristics of large noise interference and the difficulty in analyzing the vibration monitoring signal of CRDM, this paper proposes a noise reduction method for the vibration signal of CRDM based on complete ensemble empirical mode decomposition with adaptive amplitude correction noise and spectral kurtosis (CEEMDAACN-SK), which can deeply reduce the vibration signal of CRDM. Firstly, the proposed CEEMDAACN algorithm is used to decompose the vibration signal of CRDM to obtain multiple intrinsic mode functions (IMF). Then, the spectral kurtosis of each IMF component is analyzed to obtain the spectral kurtosis map of each IMF component, which is compared with the spectral kurtosis map of the original signal. Finally, the denoising reconstruction of the signal is carried out to obtain the final denoising signal. Through experimental analysis, the performance of the proposed CEEMDAACN-SK denoising algorithm is better than the complete ensemble empirical mode decomposition with adaptive noise and spectral kurtosis (CEEMDAN-SK) algorithm in terms of results. The method proposed in this paper can be applied not only to the vibration signal noise reduction of CRDM, but also to other equipment and fields, such as nuclear power main circulation pump and the chemical industry.

Caputo-Fabrizio fractional order derivative stochastic resonance enhanced by ADOF and its application in fault diagnosis of wind turbine drivetrain

Fault diagnosis of wind turbine drivetrains is vital to maintain the reliability of wind turbines and stochastic resonance (SR) is regarded as a powerful method to amplify the fault-induced weak characteristics of vibration signals. However, without considering the high dependence between current and previous values of vibration signals, integer-order SR may not be effective enough to amplify the weak characteristics. Moreover, non-Gaussian noise originating from severe working condition generally reduce the efficiency of amplification. To address these issues, a Caputo-Fabrizio fractional order derivative (CF) SR improved by ascending density outlier factor (ADOF) for fault diagnosis of wind turbine drivetrains is proposed in this paper. First, ADOF is constructed to remove the large impulsive noise of vibration signals. Second, CF is discretized into the second-order SR model, then the weak characteristics of vibration signal are amplified by the established CF second-order SR model (CF-SR). The effectiveness is validated by a simulation and two experimentations based on two real vibration signals collected from the key components of wind turbines. Compared with traditional diagnosis methods such as deconvolution and Kurtogram, the proposed method is superior for fault diagnosis of wind turbines working in severe environments.

A Deep Learning Model with Signal Decomposition and Informer Network for Equipment Vibration Trend Prediction.

Accurate equipment operation trend prediction plays an important role in ensuring the safe operation of equipment and reducing maintenance costs. Therefore, monitoring the equipment vibration and predicting the time series of the vibration trend is one of the effective means to prevent equipment failures. In order to reduce the error of equipment operation trend prediction, this paper proposes a method for equipment operation trend prediction based on a combination of signal decomposition and an Informer prediction model. Aiming at the problem of high noise in vibration signals, which makes it difficult to obtain intrinsic characteristics when directly using raw data for prediction, the original signal is decomposed once using the variational mode decomposition (VMD) algorithm optimized by the improved sparrow search algorithm (ISSA) to obtain the intrinsic mode function (IMF) for different frequencies and calculate the fuzzy entropy. The improved adaptive white noise complete set empirical mode decomposition (ICEEMDAN) is used to decompose the components with the largest fuzzy entropy to obtain a series of intrinsic mode components, fully combining the advantages of the Informer model in processing long time series, and predict equipment operation trend data. Input all subsequences into the Informer model and reconstruct the results to obtain the predicted results. The experimental results indicate that the proposed method can effectively improve the accuracy of equipment operation trend prediction compared to other models.

Application of IPSO-MCKD-IVMD-CAF in the compound fault diagnosis of rolling bearing

The vibration signals of rolling bearings are mixed with a variety of noises and a variety of faults which are coupled to each other, and which make the fault frequencies interfere with each other. This in turn leads to difficulty of accurate extraction of the features of composite faults. Therefore, it is necessary not only to suppress the noise of vibration signals but also to separate and extract the features of compound faults. A compound fault feature extraction method based on improved particle swarm optimization (IPSO) algorithm optimized maximum correlation kurtosis deconvolution (MCKD), improved variational mode decomposition (IVMD) and cyclic autocorrelation function (CAF), named IPSO-MCKD-IVMD-CAF is proposed. Firstly, the parameters of MCKD are selected adaptively by the IPSO, and the original signal is pre-processed by the parameterized MCKD. Secondly, the IVMD is utilized to decompose the signal, several intrinsic mode functions (IMFs) are obtained, the kurtosis and correlation coefficients of each IMF component are calculated. Finally, appropriate IMF components are selected by kurtosis and correlation coefficient to be superimposed into the reconstructed signals, and the reconstructed signals are analyzed by CAF. The proposed method has been successfully applied in simulation and measured vibration signal analysis, and the features of compound faults can be separated and extracted with high accuracy. The IPSO algorithm shortens the optimization time, and the IVMD solves the parameter selection problem. The analysis results of simulation case and measured data show the advantages of the proposed method IPSO-MCKD-IVMD-CAF. Compared to existing methods, the proposed method not only increases the accuracy of compound fault feature extraction, but also has a very good performance on the separation and extraction of compound fault features. More importantly, the improved algorithm reduces the computational complexity.

Noise reduction method for non-stationary noise of vibration signal of shearer mechanical transmission system

Aiming at the problem that vibration signal of mechanical transmission system of shearer is vulnerable to be interferenced by non-stationary strong noise, a combined method with variational mode decomposition and recursive least squares (VMD-RLS) is proposed for non-stationary noise reduction. Firstly, the vibration signal is decomposed by VMD and the number of decomposition layers is determined according to the principle of minimum energy difference. Then, a dominant criterion of frequency-domain characteristic amplitude is proposed to select invalid intrinsic mode components (IMF, Intrinsic Mode Function). The proportion of the frequency-domain characteristic amplitude of each order IMF component for VMD is calculated. The invalid intrinsic mode functions are selected for reconstruction, and the noise signal after VMD decomposition is obtained. Finally, the noise signal decomposed by VMD is used as reference noise signal of RLS adaptive filtering algorithm to carry out adaptive noise reduction, and the final vibration signal after noise reduction is obtained. The wavelet de-noising method, VMD de-noising method, and VMD-RLS de-noising method are used to carry out noise reduction comparative experiment on the vibration signal of shearer. The experiment results indicate that the noise reduction method of VMD-RLS is better than VMD and wavelet de-noising method. The noise reduction method of VMD-RLS can effectively extract vibration signals from non-stationary strong noise, and provide help for the subsequent fault diagnosis and analysis of shearer mechanical transmission system.

OSESgram: Data-Aided Method for Selection of Informative Frequency Bands for Bearing Fault Diagnosis

Since bearings are key components of rotatory machines and they are prone to have faults, bearing fault diagnosis has received much attention. Because vibration signals collected from the casing of a machine contain sufficient fault information, bandpass filtering-based envelope demodulation on vibration signals has been a standard approach for bearing fault diagnosis. Before envelope demodulation, selection of informative frequency bands (IFBs) for bandpass filtering to enhance fault signatures is of vital importance to achieve a good bearing fault diagnosis performance. Even though a fast kurtogram as a classic approach for IFB selection has been widely studied and many variants were proposed, they are still affected by interference caused by impulsive noise in the time domain and discrete frequencies in the frequency domain. To solve this tough problem, this paper proposes an OSESgram for IFB selection. The core of the proposed OSESgram is that healthy and faulty vibration signals are formulated as a convex optimization problem to aid IFB selection, i.e., a data-driven optimized square envelope spectrum (OSES) rather than a square envelope or a square envelope spectrum is quantified to select an IFB. Since the OSES is obtained in a data-driven way, the proposed OSESgram is promising to be more robust to interferential noise in vibration signals, especially under the existence of random impulsive noise. Benchmark and industrial bearing vibration datasets are used to validate the effectiveness and superiority of the proposed OSESgram. Experimental results demonstrated that the proposed OSESgram is more effective and robust in indicating an IFB than the fast kurtogram and an improved kurtogram.

Initial Fault Diagnosis of Rolling Bearing Based on Second-Order Cyclic Autocorrelation and DCAE Combined With Transfer Learning

Under actual working conditions, the initial faults of rolling bearings are difficult to be predicted effectively because of the lack of prior fault knowledge, weak fault information, and strong noise interference. How to enhance the fault characteristics while removing the vibration signal noise of rolling bearings has become the key challenge to the field of early fault diagnosis of rolling bearings. Based on the above problems, a rolling bearing initial fault diagnosis model based on second-order cyclic autocorrelation (Soca) and deep auto-encoder (AE) combined with transfer learning (TL) was proposed to improve the overall performance of rolling bearing fault identification. First, the Soca is used to estimate the second-order cyclic statistics of the vibration signal, which can better highlight the periodic pulse characteristics of rolling bearings and suppress random noise to a certain extent. After that, the dataset containing the valuable health state information of the rolling bearing is input into the denoising and contraction auto-encoder (DCAE). The advantages of denoising AEs (DAEs) and contraction AEs are fully utilized to extract the initial fault characteristics of rolling bearings. Finally, the balanced distribution adaptation (BDA) is used to reduce the distribution difference and class spacing of the transferable features extracted from the original datasets and the target datasets, and the initial fault diagnosis of the target rolling bearing is diagnosed according to the feature knowledge of the source rolling bearing. In three experimental datasets, six TL experiments were carried out to verify that the Soca-Denoising auto-encoder, Contraction auto-encoder and Transfer learning (DCT) model can effectively enhance the initial fault characteristics of rolling bearings, and has higher diagnostic accuracy and robustness compared with the existing initial fault diagnosis methods of rolling bearings. This method has a good application prospect in the field of early fault diagnosis of rolling bearings.

Using Windowed Interpolation Method to Suppress Power Frequency Noise of Vibration Signal

Power frequency noise interference is widespread. In order to suppress power frequency noise coupled into vibration signal, windowed interpolating method was studied. Firstly, three typical window functions were compared, including rectangular window, Hanning window and six-term cosine window. Secondly, the numerical simulation method was used to analyze the influence of window function on the interpolation results of power frequency noise, finding that different window function should be chose when the spectrum spacing between power frequency noise and useful components of vibration signal is different. Finally, the windowed interpolation method was applied to actual engineering vibration signal to suppress its power frequency noise, with the result that the interpolation result of vibration signal is the most accurate with Hanning window. Numerical simulation and engineering application show that the windowed interpolation method is effective for power frequency noise suppression.

OrbitNet: A new CNN model for automatic fault diagnostics of turbomachines

Unplanned outage due to faults in a high-fidelity turbomachine such as steam turbine and centrifugal compressor often results in the reduced reliability and productivity of a factory while increasing its maintenance costs. Shaft orbit images generated from turbomachine vibration signals have been used to diagnose component faults. However, the existing methods were developed mostly by either using features extracted from orbits or utilizing simulation data which may produce inaccurate results in practical applications due to system complexity and data uncertainties. This paper presents a novel deep learning convolution neural network methodology for accurately automatic diagnostics of multiple faults in general rotating machines by adeptly integrating advanced signal processing with orbit images augmentation, considering the high non-linearity and uncertainty of sensed vibration signals. Environmental noise in vibration signals are filtered through the integration of multiresolution discrete wavelet packet transform and Bayesian hypothesis testing-based automatic thresholding. Shaft orbit images generated from the cleansed vibration data are augmented to increase their representativity and generalization. A novel multi-layer convolutional neural network model, OrbitNet, is specially designed to improve its generality and robustness while avoid possible overfitting in fault identification of various turbomachines. The proposed model retains the pattern information in the axis trajectory to the greatest extent, with the ability of accurately capturing features of various faults in different turbomachines. A generic implementation procedure is proposed for automatic fault diagnosis of rotating machinery based on the presented methodology. A comparison study is conducted to demonstrate the effectiveness and feasibility of the proposed methodology by using the sensed vibration signals collected from three real-world centrifugal compressors, two steam turbines and one generator with four different fault modes including imbalance, friction, misalignment and oil whirl.

Multisource Fault Signal Separation of Rotating Machinery Based on Wavelet Packet and Fast Independent Component Analysis

The vibration signal of rotating machinery compound faults acquired in actual fields has the characteristics of complex noise sources, the strong background noise, and the nonlinearity, causing the traditional blind source separation algorithm not be suitable for the blind separation of rotating machinery coupling fault. According to these problems, an extraction method of multisource fault signals based on wavelet packet analysis (WPA) and fast independent component analysis (FastICA) was proposed. Firstly, according to the characteristic of the vibration signal of rotating machinery, an effective denoising method of wavelet packet based on average threshold is presented and described to reduce the vibration signal noise. In the method, the thresholds of every node of the best wavelet packet basis are acquired and averaged, and then the average value is used as a global threshold to quantize the decomposition coefficient of every node. Secondly, the mixed signals were separated by using the improved FastICA algorithm. Finally, the results of simulations and real rotating machinery vibration signals analysis show that the method can extract the rotating machinery fault characteristics, verifying the effectiveness of the proposed algorithm.

Domain adaptation-based deep feature learning method with a mixture of distance measures for bearing fault diagnosis

The success of deep learning-based bearing fault diagnosis depends on (a) training data and test data following the same data distribution; (b) a mass of labeled fault data being available. However, the working conditions of the bearings is changing, which leads to a difference in data distribution, and it is very laborious to obtain a mass of labeled data with fault information. To address these issues, a domain adaptation-based deep feature learning method with a mixture of distance measures for bearing fault diagnosis is proposed. First, the noise in vibration signal is filtered by wavelet packet decomposition and reconstruction. Then, frequency slice wavelet transform is used to transform the reconstructed signal into a two-dimensional time–frequency image. Furthermore, a domain-adaptative deep neural network based on ResNet50 is used for bearing fault diagnosis under different working conditions. A mixture of distance measures is used to minimize the distribution discrepancy between source and target domains. The bearing datasets provided by Case Western Reserve University and Paderborn University verify the effectiveness of the proposed method.

An Enhanced Intrinsic Time-Scale Decomposition Method Based on Adaptive Lévy Noise and Its Application in Bearing Fault Diagnosis

When early failures in rolling bearings occur, we need to be able to extract weak fault characteristic frequencies under the influence of strong noise and then perform fault diagnosis. Therefore, a new method is proposed: complete ensemble intrinsic time-scale decomposition with adaptive Lévy noise (CEITDALN). This method solves the problem of the traditional complete ensemble intrinsic time-scale decomposition with adaptive noise (CEITDAN) method not being able to filter nonwhite noise in measured vibration signal noise. Therefore, in the method proposed in this paper, a noise model in the form of parameter-adjusted noise is used to replace traditional white noise. We used an optimization algorithm to adaptively adjust the model parameters, reducing the impact of nonwhite noise on the feature frequency extraction. The experimental results for the simulation and vibration signals of rolling bearings showed that the CEITDALN method could extract weak fault features more effectively than traditional methods.

Multi-task learning of classification and denoising (MLCD) for noise-robust rotor system diagnosis

Deep learning-based research has drawn much attention in the field of fault diagnosis of various mechanical systems due to its powerful performance. In deep learning-based methods, signals become the input for a deep learning algorithm. However, the performance of an algorithm can be diminished if there is significant noise in the data. To address the noise issue, this paper proposes a fault diagnosis method called multi-task learning of classification and denoising (MLCD). The proposed method is designed to make a fault diagnosis algorithm robust against the noise in vibration signals by learning the denoising task simultaneously with the classification. Given a noisy input, MLCD can improve test accuracy by implementing denoising as an auxiliary task, using hyperparameters chosen by Bayesian optimization. To validate the proposed MLCD method, it is integrated with the two most commonly used deep learning algorithms: long short-term memory and one-dimensional convolutional neural network. For a case study, these algorithms are tested to classify the states of a rotor testbed data. The results show that the proposed MLCD method extracts noise-robust and meaningful features; ultimately, this improves the fault diagnosis performance.

A novel patch-matching 2D denoising method for fault diagnosis of roller bearings

The vibration signal of roller bearings contains important information, but the strong background noise makes fault diagnosis difficult. In this paper, inspired by the idea of a block-matching 3D algorithm, using local and nonlocal correlation of vibration signal, a patch-matching 2D (PM2D) denoising method is proposed for the first time to suppress noise in vibration signals. The proposed denoising method constructs similarity matrices of component modules, which are used for threshold processing to determine the coefficients of the 2D discrete cosine transform, so as to achieve optimal denoising performance. Then, empirical mode decomposition and envelope analysis are employed to perform fault diagnosis. The proposed PM2D denoising method and fault diagnosis strategies are applied to both simulated and measured signals. A comparison study shows the superiority of the proposed method over the other existing denoising methods.

Application of weighted contribution rate of nonlinear output frequency response functions to rotor rub-impact

Abstract The rub-impact between a rotor and a stator is a common fault of the rotor system. Various methods are available to detect the rub impact using vibration signals. However, the noise in vibration signals causes inconvenience when extracting characteristic signals. Nonlinear output frequency response functions (NOFRFs) comprise an effective method for using input and output signals to detect damage of a structure or system. A new evaluation method named the NOFRFs weighted contribution rate has been proposed in this paper. This method is an extension of traditional NOFRFs-based indexes and attempt to obtain a new index with higher recognition of faults. As a new index extracted by this method, the second-order optimal weighted contribution rate of the NOFRFs optimally amplify the extracted fault features. The characteristic signal after amplification is significantly enhanced. The new index is applied to the evaluation of a rub-impact fault. The simulation and experimental results demonstrate the sensitivity of the new index to rub-impact faults. Especially in the initial stage of the occurrence of rub-impact, the new index shows better superiority. The findings demonstrated that the second order optimal weighted contribution rate has a high application value in the fault diagnosis of rub-impact.

Multidimensional denoising of rotating machine based on tensor factorization

Abstract A multidimensional de-noising technique of rotating machinery based on the tensor factorization is developed for solving multidimensional signal filtering. The vibration signal is first formulated as a 3-way tensor using temporal signal, frequency and channel information. Tensor model is then factorized via the truncated high-order singular value decomposition. The L-curve criterion is adopted to determine the truncation parameters used in the tensor factorization. The performance of the developed technique in detecting faults of rotating machinery has been thoroughly evaluated through simulated signals compared with three traditional techniques. The presented approach is then applied to reduce noise in vibration signal collected on bearing and gear test-rigs. Experimental results showed that it can well remove noise and retain the fine signatures as much as possible. This presented tensor based multidimensional signal filtering will broaden the view in dealing with heterogeneous and multiaspect data in an age of big data.

Fault Detection of Broken Rotor Bar Using an Improved form of Hilbert–Huang Transform

Signal processing is an integral part in signal-based fault diagnosis of rotary machinery. Signal processing converts the raw data into useful features to make the diagnostic operations. These features should be independent from the normal working conditions of the machine and the external noise. The extracted features should be sensitive only to faults in the machine. Therefore, applying more efficient processing techniques in order to achieve more useful features that bring faster and more accurate fault detection procedure has attracted the attention of researchers. This paper attempts to improve Hilbert–Huang transform (HHT) using wavelet packet transform (WPT) as a preprocessor instead of ensemble empirical mode decomposition (EEMD) to decompose the signal into narrow frequency bands and extract instantaneous frequency and compares the efficiency of the proposed method named “wavelet packet-based Hilbert transform (WPHT)” with the HHT in the extraction of broken rotor bar frequency components from vibration signals. These methods are tested on vibration signals of an electro-pump experimental setup. Moreover, this project applies wavelet packet de-noising to remove the noise of vibration signal before applying both methods mentioned and thereby achieves more useful features from vibration signals for the next stages of diagnosis procedure. The comparison of Hilbert transform amplitude spectrum and the values and numbers of detected instantaneous frequencies using HHT and WPHT techniques indicates the superiority of the WPHT technique to detect fault-related frequencies as an improved form of HHT.

Rolling Element Bearing Condition Monitoring Based on Vibration Analysis Using Statistical Parameters of Discrete Wavelet Coefficients and Neural Networks

There are several techniques that can be used to determine the condition of a rolling element bearing. In this paper, vibration analysis is used to conduct fault diagnosis of a bearing. Vibration signal noise was eliminated using hard thresholding wavelet analysis. The best mother wavelet for the denoising process was selected using the minimum Shannon entropy criterion. Statistical parameters and other signal properties such as energy and entropy are powerful tools for analyzing vibration signals. These features were calculated in the time and wavelet domains and applied to Artificial Neural Networks (ANNs) as the feature vector to classify the condition of a bearing into one healthy and three faulty conditions. The ANN parameters were separately optimized using three optimization algorithms. The comparison of the results shows that if the ANN parameters are properly optimized, the statistical parameters in the time-frequency domain can optimize accuracy.

Rolling Element Bearing Condition Monitoring Based on Vibration Analysis Using Statistical Parameters of Discrete Wavelet Coefficients and Neural Networks

There are several techniques that can be used to determine the condition of a rolling element bearing. In this paper, vibration analysis is used to conduct fault diagnosis of a bearing. Vibration signal noise was eliminated using hard thresholding wavelet analysis. The best mother wavelet for the denoising process was selected using the minimum Shannon entropy criterion. Statistical parameters and other signal properties such as energy and entropy are powerful tools for analyzing vibration signals. These features were calculated in the time and wavelet domains and applied to Artificial Neural Networks (ANNs) as the feature vector to classify the condition of a bearing into one healthy and three faulty conditions. The ANN parameters were separately optimized using three optimization algorithms. The comparison of the results shows that if the ANN parameters are properly optimized, the statistical parameters in the time-frequency domain can optimize accuracy.

A method of eliminating the vibration signal noise of hydropower unit based on NA-MEMD and approximate entropy

A new vibration signal denoising method of hydropower unit based on noise-assisted multivariate empirical mode decomposition (NA-MEMD) and approximate entropy is proposed. Firstly, the NA-MEMD is used to decompose the signal into a number of intrinsic mode functions. Then, the approximate entropy of each component is computed. According to a preset threshold of approximate entropy, these components are reconstructed to denoise vibration signal of hydropower unit. The analysis results of simulation signal and real-world signal show that the proposed method is adaptive and has a good denoising performance. It is very suitable for online denoising of hydropower unit's vibration signal.