Ensemble Local Mean Decomposition Research Articles

Detection of ballastless track interlayer gap based on vehicle’s multivariate dynamic response and deep learning

To adapt to the rapid detection of interlayer damage in ballastless track structures of high-speed railways, a cement asphalt mortar (CAM) gap localization and damage degree classification scheme based on multivariate data fusion and deep learning is proposed. Based on vertical axle box acceleration (VABA) and vertical wheel-rail force (VWRF) data, the variation patterns of multi-dynamic response data of vehicles under the edge type and internal type interlayer gaps are analyzed. An improved ensemble local mean decomposition algorithm (IELMD) is designed to achieve data denoising and preliminary enhancement preprocessing of weak damage features for VABA and VWRF under interlayer gaps. The measured and simulated VABA and VWRF constituted multiple dynamic response data sets for interlayer gap detection. An interlayer gap detection model: DTA-Tcnformer is constructed, integrating a dual temporal convolutional network with an attention mechanism, transformer architecture, and gate control mechanism. Multiple dynamic response data are input into the model to locate and classify 24 cases of the two interlayer gap types. The influence of damage cases and vehicle speeds on the location effect is analyzed. The learning ability of the proposed model on the gap features is visualized. The misidentification and omission of the gap cases are calculated, and the comprehensive recognition accuracy is 92.48 %. The recognition performance of the proposed model is compared and verified.

Fault diagnosis of anti-friction bearings based on Bi-dimensional ensemble local mean decomposition and optimized dynamic least square support vector machine

In order to ensure the normal operation of rotating equipment, it is very important to quickly and efficiently diagnose the faults of anti-friction bearings. Hereto, fault diagnosis of anti-friction bearings based on Bi-dimensional ensemble local mean decomposition and optimized dynamic least square support vector machine (LSSVM) is presented in this paper. Bi-dimensional ensemble local mean decomposition, an extension of ensemble local mean decomposition from one-dimensional signal processing to Bi-dimensional signal processing, is used to extract the features of anti-friction bearings. Moreover, an optimized dynamic LSSVM is used to fault diagnosis of anti-friction bearings. The experimental results show that Bi-dimensional ensemble local mean decomposition is superior to Bi-dimensional local mean decomposition, optimized dynamic LSSVM is superior to traditional LSSVM, and the proposed Bi-dimensional ensemble local mean decomposition and optimized dynamic LSSVM method is effective for fault diagnosis of anti-friction bearings.

Rolling bearing fault diagnosis method based on ELMD hybrid feature extraction and wavelet neural network

Rolling bearings are the most important components in the transmission system of coal mining machinery, and their operating condition significantly impacts the entire mechanical and electrical equipment. Therefore, the fault diagnosis of rolling bearing can effectively ensure the operation reliability of equipment. Given the strong noise, coal impact, and other interference, the vibration signal of the rolling bearing cannot be effectively decomposed, and the fault identification efficiency is low. According to the method based on vibration analysis, this article proposes a rolling bearing fault diagnosis method based on ensemble local mean decomposition (ELMD) hybrid feature extraction and wavelet neural network. ELMD is used to solve the problem of modal aliasing in local mean decomposition (LMD), which can improve the efficiency of LMD. Quantitatively extracting the mixed features of each component and introducing a wavelet neural network for fault type recognition. The experimental results demonstrate that the proposed method has a high accuracy in fault recognition and is an effective fault diagnosis method.

A Multi-Indicator Fusion-Based Approach for Fault Feature Selection and Classification of Rolling Bearings

Concerning the problems of harrowing extraction and poor classification accuracy of fault features in rolling bearing vibration signals, a fault feature selection and classification method based on multi-indicator fusion is proposed. Firstly, the original signal is decomposed through the improved complementary ensemble local mean decomposition (CELMD) method into several physically meaningful product functions (PF) and single residual components; then, the three indicators of kurtosis, correlation coefficient, and K-L divergence are combined to extract the most suitable PF components for signal reconstruction. Ultimately, the reconstructed signal's multi-domain characteristics and entropy value features are retrieved and fed into the LightGBM classifier for classification in order to achieve an intelligent diagnosis of rolling bearing problems. The statistical results demonstrate that the proposed method can efficiently identify the functional PF components and has notable benefits in extracting features from diverse experimental data sets and detecting faults.

A signal-filtering and feature-enhancement method based on ensemble local mean decomposition and adaptive morphological filtering

The bearing fault signals from the spindle motors of computer numerical control machines are complex and non-linear due to being coupled to multiple subsystems. The complexity of industrial signals, with increased industrial noise, and the difference in fault features in different life cycles and different individual signals bring great challenges for fault feature extraction. In this paper, a signal-filtering and feature-enhancement method based on an ensemble local mean decomposition and adaptive morphological filtering (ELMD-AMF) method is proposed. First, the original vibration signal of the bearing is reconstructed by ELMD to reducing interference from background noise. Next, an improved feature-enhancement process based on AMF is constructed, a particle swarm optimization with maximum-weighted spectral kurtosis as an optimization objective is used to adaptively construct the size of the structural element, and a morphology hat product operator one is adapted to extract the periodic impulse features. Finally, the effectiveness of the method is proved by using an actual three-phase induction motor matched with an NTN ceramic bearing and a FAG metal bearing, respectively. Further, compared with minimum entropy deconvolution and fast kurtogram methods, the result proves that the proposed method has better performance for both early-failure and late-failure scenarios under real-world engineering conditions.

Ensemble robust local mean decomposition integrated with random forest for short-term significant wave height forecasting

A robust short-term significant wave height (Hs) modelling framework based on an ensemble local mean decomposition method integrated with random forest (i.e., En-RLMD-RF) is developed. The robust local mean decomposition (RLMD) decomposed the Hs data series into three subseries; amplitude modulation, frequency modulation and the low-frequency product function (PFs). The partial autocorrelation function was employed to determine the correlation-based significant predictor signals between the PFs at t0 and t1. Then the statistically significant PFs were incorporated into the random forest (RF) to construct the RLMD-RF model. The RLMD-RF based forecasted PFs were used again in the RF model as input predictors resulting in an ensemble-based RLMD-RF (i.e., En-RLMD-RF) model for forecasting short-term Hs. The En-RLMD-RF model is validated and compared with RF, extreme learning machine (ELM) and multiple linear regression (MLR) models and their hybrids RLMD-RF, RLMD-ELM, RLMD-MLR, En-RLMD-ELM and En-RLMD-MLR counterparts using a set of performance metrics. The results demonstrated that the En-RLMD-RF model generates better forecasting accuracy against the benchmarking models. This study is beneficial for the application and optimization of more clean energy resources worldwide for sustained energy generation.

A Bias Drift Suppression Method Based on ICELMD and ARMA-KF for MEMS Gyros.

The applications of Micro-Electro-Mechanical-System (MEMS) gyros in inertial navigation system is gradually increasing. However, the random drift of gyro deteriorates the system performance which restricting the applications of high precision. We propose a bias drift compensation model based on two-fold Interpolated Complementary Ensemble Local Mean Decomposition (ICELMD) and autoregressive moving average-Kalman filtering (ARMA-KF). We modify CELMD into ICELMD, which is less complicated and overcomes the endpoint effect. Further, the ICELMD is combined with ARMA-KF to separate and simplify the preprocessed signal, resulting improved denoising performance. In the model, the abnormal noise is removed in preprocess by 2σ criterion with ICELMD. Then, continuous mean square error (CMSE) and Permutation Entropy (PE) are both applied to categorize the preprocessed signal into noise, mixed and useful components. After abandon the noise components and denoise the mixed components by ARMA-KF, we rebuild the noise suppression signal of MEMS gyro. Experiments are carried out to validate the proposed algorithm. The angle random walk of gyro decreases from 2.4156∘/h to 0.0487∘/h, the zero bias instability lowered from 0.3753∘/h to 0.0509∘/h. Further, the standard deviation and the variance are greatly reduced, indicating that the proposed method has better suppression effect, stability and adaptability.

Fault Frequency Identification of Rolling Bearing Using Reinforced Ensemble Local Mean Decomposition

The vibration signal of rolling bearing exhibits the characteristics of energy attenuation and complex time-varying modulation caused by the transmission with multiple interfaces and complex paths. In view of this, strong ambient noise easily masks faulty signs of rolling bearings, resulting in inaccurate identification or even totally missing the real fault frequencies. To overcome this problem, we propose a reinforced ensemble local mean decomposition method to capture and screen the essential faulty frequencies of rolling bearing, further boosting fault diagnosis accuracy. Firstly, the vibration signal is decomposed into a series of preliminary features through ensemble local mean decomposition, and then the frequency components above the average level are energy-enhanced. In this way, principal frequency components related to rolling bearing failure can be identified with the fast spectral kurtosis algorithm. Finally, the efficacy of the proposed approach is verified through both a benchmark case and a practical platform. The results show that the selected fault characteristic components are accurate, and the identification and diagnosis of rolling bearing status are improved. Especially for the signals with strong noise, the proposed method still could accurately diagnose fault frequency.

Genetic algorithms to determine the optimal parameters of an ensemble local mean decomposition

An optimization method for an ensemble local mean decomposition (ELMD) parameters selection using genetic algorithms is proposed. The execution of this technique depends heavily on the correct choice of the parameters of its model as pointed out in previous works. The effectiveness of the proposed method was evaluated using synthetic signals, discussed by several authors. The resulting algorithm obtained similar results to OELMD, but with an 82% reduction in processing time. Actual vibration signals were also analysed, presenting satisfactory results.

Non-contact harmonic magnetic field detection for parallel steel pipeline localization and defects recognition

A novel detection method and targets signal extraction algorithm for parallel steel pipeline localization and defects recognition based on harmonic magnetic field are proposed. An orthogonal vector differential focusing array detector was designed, which radiates high-frequency (HF) and ultra-low-frequency (ULF) electromagnetic waves to pipeline based on the principle of Frequency Modulated Continuous Wave (FMCW), and records the signals in parallel using high-sensitivity Tunnel Magneto Resistance (TMR) array sensors. The joint signal pre-processing algorithm of Combination Morphological Pulse Elimination (CMPE) and Quantum Genetic Gaussian Potential Stochastic Resonance (QGGPSR) is proposed to improve signal quality by adaptively eliminating random pulse and noise. The time–frequency analysis algorithm Complementary Ensemble Local Mean Decomposition (CELMD) is used to decompose the pipeline defects signal in detail, and the appropriate components are selected and fused into a correlation array signal cluster. Finally, the targets signal are separated by Limited-memory BFGS Independent Component Analysis (L-BFGS-ICA) based on machine learning optimization to enhance their legibility. The magnetic dipole harmonic simulation model and the 20# steel parallel pipeline experimental platform are established, and compared with the passive detection method to verify the effectiveness and practicality of the proposed method. The results show that this method can accurately locate parallel pipelines and detect different kinds of defects in noisy environment, which is of great significance for the application of non-contact Non-Destructive Testing (NDT) in practical engineering.

Application of adaptive complementary ensemble local mean decomposition in underwater acoustic signal processing

The aim of this study is to develop a signal decomposition method that not only overcomes the shortcomings of existing decomposition methods, but also can be further applied to complex underwater acoustic signal processing. Therefore, a novel Adaptive Complementary Ensemble Local Mean Decomposition (ACELMD) method was proposed. A simulated signal and real-world underwater acoustic signals from three marine mammals (white-sided dolphin, long-finned pilot whale, and harp seal) were employed to evaluate the decomposition performance, reliability, and practicality of the proposed ACELMD. Also, the decomposition results of the same simulated signal using ACELMD, Local Mean Decomposition (LMD), Ensemble Local Mean Decomposition (ELMD), and Complementary Ensemble Local Mean Decomposition (CELMD) methods were compared. All the results demonstrate the excellent decomposition performance of the proposed method. Moreover, compared with the other three methods, ACELMD effectively reduces the modal aliasing in the decomposition results, reduces the number of LMD executions, and is more inclusive of the white noise amplitude, all of which indicate its great potential in practical applications.

Analyzing chatter vibration during turning on computer numerical control lathe using ensemble local mean decomposition and probabilistic approach

Chatter vibration is an undesired and indispensable phenomenon in turning operation, which cannot be completely avoided. However, it can be suppressed by early identification and with the proper choice of input turning parameters. The key issue of chatter detection is to process the acquired signals and extract the features pertaining to it. In the present work, a methodology has been proposed for exploring tool chatter features in the incipient stage during turning on lathe. Chatter signals generated during the turning of Al 6061-T6 have been acquired using a microphone. A stability lobe diagram has been plotted to access the stability regime. Further, in order to study the effect of feed rate on stability, the recorded signals have been processed using a local mean decomposition signal processing technique, followed by the selection of dominating product functions using the Fourier transform. The decomposed signals have been used to evaluate the new output parameter, that is, chatter index. Further, the Nakagami probability distribution has been used to ascertain stability region (threshold). From the experimental validation, it has been inferred that cutting combinations obtained from the Nakagami probability distribution are significant and capable of limiting chatter vibrations. The present methodology will serve as guidelines to the researchers and machinist for the identification of tool chatter in the incipient stage, explore its severity, and finally suppress it with the proper selection of input turning parameters.

A fault pattern recognition method for rolling bearing based on CELMDAN and fuzzy entropy

The vibration signal of rolling bearing often has the characteristics of strong noise, nonlinearity and non-stationary, so the accurate fault feature information cannot be obtained directly from the measured vibration signal. Therefore, a fault pattern recognition method for rolling bearing based on complete ensemble local mean decomposition with adaptive noise (CELMDAN) and fuzzy entropy is deeply studied. Firstly, the reason of modal aliasing existing in local mean decomposition (LMD) method is explained. According to the previous methods for modal aliasing processed in other methods, CELMDAN method is proposed. The experiment proves that the proposed CELMDAN method can better handle the vibration signals with nonlinear and non-stationary. Then, the principle and properties of the fuzzy entropy are introduced in detail, and the fault feature of rolling bearing can be reflected. Finally, extreme learning machine (ELM) is introduced as the pattern recognition method based on the effective fault feature of rolling bearing. Combined with the verification of experimental signal, it is proved that the proposed method can extract the fault features of rolling bearing accurately and effectively, and the fault pattern recognition of rolling bearing can be realized.

Research on gear fault degree recognition method based on multi sensor fusion

The fatal fault of gear will inevitably experience the evolution of different fault degrees, and the accurate recognition of the fault degree of gear has more practical significance for predictive maintenance and efficient operation. A gear fault degree recognition method based on multi sensor fusion is proposed. Ensemble local mean decomposition (ELMD) is used to decompose the vibration signal of gear, which can eliminate the aliasing effect of LMD. Then, the fault feature is extracted by using envelope spectrum information entropy and time-domain kurtosis. Based on the initial recognition of the sub evidence formed by a signal sensor based on wavelet neural network (WNN), the basic belief function assignment and fusion methods of Dempster-Shafer (D-S) evidence theory are determined, and the accurate recognition based on multi sensor fusion for different fault degree of gear is realized, which is verified by experiments.

A Novel Flash P2P Network Traffic Prediction Algorithm based on ELMD and Garch

To improve the quality of service and network performance for the Flash P2P video-on-demand, the prediction Flash P2P network traffic flow is beneficial for the control of the network video traffic. In this paper, a novel prediction algorithm to forecast the traffic rate of Flash P2P video is proposed. This algorithm is based on the combination of the ensemble local mean decomposition (ELMD) and the generalized autoregressive conditional heteroscedasticity (GARCH). The ELMD is used to decompose the original long-related flow into the summation of the short-related flow. Then, the GRACH is utilized to predict the short-related flow. The developed algorithm is tested in a university’s campus network. The predicted results show that our proposed method can further achieve higher accuracy than those obtained by existing algorithms, such as GARCH and Local Mean Decomposition and Generalized AutoRegressive Conditional Heteroskedasticity (LMD-GARCH) while keeping lower computational complexity.

A Hybrid Time-Frequency Analysis Method for Railway Rolling-Element Bearing Fault Diagnosis

The health condition of rolling-element bearings is important for machine performance and operating safety. Due to external interferences, the impulse-related fault information is always buried in the raw vibration signal. To solve this problem, a hybrid time-frequency analysis method combining ensemble local mean decomposition (ELMD) and the Teager-Kaiser energy operator (TKEO) is proposed for the fault diagnosis of high-speed train bearings. The ELMD method is a significant improvement over local mean decomposition (LMD) for addressing the mode-mixing problem. The TKEO method is effective for separating amplitude-modulated (AM) and frequency-modulated (FM) signals from a raw signal. But it is only valid for monocomponent AM-FM signals. The proposed time-frequency method integrates the advantages of ELMD and TKEO to detect localized defects in rolling-element bearings. First, a raw signal is decomposed into an ensemble of PFs and a residual component using ELMD. A novel sensitive parameter (SP) is introduced to select the sensitive PF that contains the most fault-related information. Subsequently, the TKEO is applied to extract both the amplitude and frequency modulations from the selected PF. The experimental results of rolling element and outer race fault signals confirmed that the proposed method could effectively recover fault information from raw signals contaminated by strong noise and other interferences.

Application of an Improved Ensemble Local Mean Decomposition Method for Gearbox Composite Fault Diagnosis

In industrial production, it is highly essential to extract faults in gearbox accurately. Specifically, in a strong noise environment, it is difficult to extract the fault features accurately. LMD (local mean decomposition) is widely used as an adaptive decomposition method in fault diagnosis. In order to improve the mode mixing of LMD, ELMD (ensemble Local Mean Decomposition) is proposed as local mode mixing exists in noisy environment, but white noise added in ELMD cannot be completely neutralized leading to the influence of increased white noise on PF (product function) component. This further leads to the increase in reconstruction errors. Therefore, this paper proposes a composite fault diagnosis method for gearboxes based on an improved ensemble local mean decomposition. The idea is to add white noise in pairs to optimize ELMD, defined as CELMD (Complementary Ensemble Local Mean Decomposition) then remove the decomposed high noise component by PE (Permutation Entropy) while applying the SG (Savitzky‐Golay) filter to smooth out the low noise in PFs. The method is applied to both simulated signal and experimental signal, which overcomes mode mixing phenomenon and reduces reconstruction error. At the same time, this method avoids the occurrence of pseudocomponents and reduces the amount of calculation. Compared with LMD, ELMD, CELMD, and CELMDAN, it shows that improved ensemble local mean decomposition method is an effective method for extracting composite fault features.

Aging State Detection of Viscoelastic Sandwich Structure Based on ELMD and Sensitive IA Spectrum Entropy

Viscoelastic sandwich structure is playing an important role in mechanical equipment, but therein viscoelastic material inevitably suffers from aging which affects structural service performance and the whole performance of equipment. Therefore, the aging state detection of viscoelastic sandwich structure based on vibration response signal is essential for monitoring the health state of structure and guaranteeing the operation safety of equipment. However, the weakness of structural vibration response variation caused by material aging make this task challenging. In this paper, a novel method based on ensemble local mean decomposition (ELMD) and sensitive IA spectrum entropy is proposed for this task. As an adaptive nonlinear and non-stationary signal processing method, ELMD is introduced to decompose the structural vibration response signal, and a series of instantaneous amplitudes (IAs) are obtained. Then, the spectrum entropies of these IAs are developed to quantitatively assess the aging state of viscoelastic sandwich structure. However, the IA spectrum entropies have different sensitivities to the aging state. Therefore, the most sensitive IA spectrum entropy is selected with a distance evaluation technique to detect the aging state of viscoelastic sandwich structure. In order to demonstrate the effectiveness of the proposed method, the experimental device of a viscoelastic sandwich structure is designed, and different structural aging states are created through the accelerated aging of viscoelastic material. The results show the outstanding performance of the proposed method.

Modified Singular Spectrum Decomposition and Its Application to Composite Fault Diagnosis of Gearboxes.

Under the strong noise environment, the composite fault signal of gearbox is weak, which makes it difficult to extract fault features. For this problem, based on noise-assisted method, we propose a novel method called Modified Singular Spectrum Decomposition (MSSD). Singular Spectrum Decomposition (SSD) has many advantages such as high decomposition precision and strong ability to restrain mode mixing, etc. However, the ability of SSD to extract a weak signal is not ideal, the decomposition results usually contain a lot of redundant noise and mode mixing caused by intermittency, which is also a troubling problem. In order to improve the decomposition efficiency and make up for the defects of SSD, the new method MSSD adds an adaptive and particular noise in every SSD decomposition stage for each trial, and in addition, whenever the input signal is decomposed to obtain an intrinsic module function (IMF), a unique residual is obtained. After multiple decomposition, the average value of the residual is used as input to the next stage, until the residual cannot continue to decompose, which means that the residual component has, at most, one extreme value. Finally, analyzing simulated signals to explain the advantages of MSSD compared to ensemble empirical mode decomposition (EEMD) and complete ensemble local mean decomposition with adaptive noise (CEEMDAN). In order to further prove the effectiveness of MSSD, this new method, MSSD, is applied to the fault diagnosis of an engineering gearbox test stand in an actual engineer project case. The final results show that MSSD can extract more fault feature information, and mode mixing has been improved and suffers less interference compared to SSD.

Application of an Improved Multipoint Optimal Minimum Entropy Deconvolution Adjusted for Gearbox Composite Fault Diagnosis.

The fault feature extraction of gearbox is difficult to achieve under complex working conditions, and this paper presents a hybrid fault diagnosis method for gearbox based on the combining product function (CPF) and multipoint optimal minimum entropy deconvolution adjusted (MOMEDA) methods. First, ensemble local mean decomposition (ELMD) is utilized to reduce the noise in original signal, and get a series of product functions (PFs), through the correlation coefficient method to remove false components and residual components. Then, multi-point kurtosis of the definition is achieved by calculating the multi-point kurtosis spectrum of each layer PF, and the fault feature period is extracted and the PFs without periodic impact are removed. After that, in order to maintain the integrity of the original signal, the PFs with the same period are recombined by the combined product function method. Finally, the different cycle interval is configured, reduce the noise through MOMEDA on the combined signal, to further extract the fault feature. The method is applied to the feature extraction of gear box composite fault to verify the feasibility of this method.