Sensor Fault Diagnosis Using Ensemble Empirical Mode Decomposition and Extreme Learning Machine

Based on ensemble empirical mode decomposition (EEMD) and the support vector machine (SVM), an algorithm used in the sensor fault detection and classification is put forward in this paper. Using this method and through EEMD, the sensor signal is decomposed into several segments, including the original signals, several intrinsic mode functions (IMFs) and the residual signals. Moreover, as features of the sensor fault, their variance, mean, entropy and the slope of the original signal are calculated in accordance with the characteristics of different fault types and the inherent physical meanings of each IMF. Subsequently, the feature vectors are inputted into the SVM, which is used to classify the detection and identification of sensor faults. Finally, the simulation results of the fault diagnosis of a carbon dioxide sensor indicate that this method may not only be effectively applied to fault diagnosis of carbon dioxide sensors but also provides a reference for that of other sensors.

A novel sensor fault diagnosis method based on Modified Ensemble Empirical Mode Decomposition and Probabilistic Neural Network

Application of the EEMD method for distinction and suppression of motion-induced noise in grounded electrical source airborne TEM system

The present study investigated the use of ensemble empirical mode decomposition (EEMD) based residual signals for the condition monitoring and fault diagnosis in the bearings. The acquired signals were decomposed using EEMD technique to obtain intrinsic mode functions (IMFs). For further processing the sensitive IMFs were selected on the basis of their similarity index with the original signal. The residual signal is then obtained by subtracting the ineffective IMFs from the effective IMFs. After that the five statistical features (namely: Root mean square value (RMS), crest factor (CF), peak-to-RMS (pk2rms), impulse factor (IF) and kurtosis (Ku)) and five entropy based features (namely: Approximate entropy (ApEn), permutation energy (PE), Sample entropy (SampEn), dispersion entropy (DE), and fuzzy entropy (FzEn)) were extracted from residual signal. The extracted features were fed to the two state-of-the-art classifiers, support vector machine (SVM) and random forest (RF) for the classification of different bearing faults. The analysis of results shows that the proposed method has potential for identifying and classifying different bearing’s faults.

To realize accurate fault diagnosis of rolling bearing under random noise, a novel fault diagnosis method based on ensemble empirical mode decomposition (EEMD) and optimized Elman_AdaBoost is proposed in this paper. First, the EEMD method is used to decompose the original vibration signal into several intrinsic mode functions (IMFs). Then, the correlation coefficient and kurtosis of IMFs are used to remove any excess ingredients, and the signal is reconstructed by using the rest of IMFs. In addition, based on the analysis of the time–frequency characteristics of the reconstructed signal, the root-mean-square value and the power spectrum center are extracted as a 2-D feature vector from each reconstructed signal and regarded as the input characteristic vector of the Elman neural network. Combined with the AdaBoost classification algorithm, the Elman_AdaBoost algorithm is improved and proposed to establish the strong classifier. Finally, based on the vibration data recorded in the bearing center of the Case Western Reserve University, three kinds of bearings with different faults and normal bearings are selected as the sample data and the training samples and test samples are constructed. Experimental results show that the proposed EEMD method and optimized Elman_AdaBoost model can effectively diagnose the rolling bearing under random noise, and has the advantages of fast speed, small error, and stable performance. Comparing with the traditional Elman_AdaBoost and single Elman neural network algorithm, this paper with fewer characteristics can get better accuracy and real-time processing performance, and presents a simple and practical resolution for the fault diagnosis of rolling bearings.

Recently, ensemble empirical mode decomposition (EEMD) method has been developed for non-linear and non-stationary signal analysis. The method can work on nature signals (non-linear and nonstationary signals) and reduce the speckle noise. With the EEMD method, the signal is decomposed into several intrinsic mode functions (IMFs) and the frequencies of IMFs are arranged in decrease order (high to low) after the EEMD processing. The scaling mode of the EEMD method is similar to wavelet transform, but the signal resolutions in different frequency domains do not decrease by down-sampling. There are a large population and a developed economy in Weihe watershed, the disasters of droughts and floods caused by the autumn precipitation (here is precipitation in September and October) less or more than normal cause great loss and serious influence. In this paper, we propose the EEMD method to decompose the autumn precipitation series in the Weihe river basin during last 50 years into several IMFs, then extract the information including in the precipitation series and get the characteristics of multi-scales. The result shows that it is well response to the autumn precipitation series in the Weihe river basin and to the abrupt climate change in late 1970s and early 1980s of last century. The response appears earlier for high time scales than for low time scales In addition, the expression of the response for high time scales is the form of variability, but it is the amplitude of variability for low time scales.

Ensemble empirical mode decomposition (EEMD) has been widely used in fault diagnosis of planetary gearboxes. However, artificially selecting the amplitude of white Gaussian noise (WGN) which is the crucial parameter of EEMD probably produces subjective errors. To alleviate this drawback, a self-adaptive EEMD method based on particle swarm optimization (PSO-EEMD) is proposed. This method takes the uniform distribution characteristic of signal extreme points as the fitness function of PSO in order to self-adaptively determine the optimal amplitude of WGN, which can improve the self-adaptability and decomposition precision of EEMD. Furthermore, utilizing the advantages of PSO-EEMD in extracting weak characteristics from noisy signals, a new fault diagnosis method of planetary gearboxes based on PSO-EEMD and frequency demodulation analysis is put forward. Simulation analysis results show that the PSO-EEMD method can self-adaptively determine the optimal amplitude of GWN according to the original signal characteristics and obtain more accurate intrinsic mode functions (IMFs) than the traditional EEMD, which illustrate the effectiveness and superiority of the proposed method. The fault diagnosis case of a planetary gearbox shows that the proposed fault diagnosis method based on PSO-EEMD and frequency demodulation analysis can effectively extract the characteristic frequency from the sun gear with a crack fault.

Various faults inevitably occur in mechanical systems and may result in unexpected failures. Hence, fault detection is critical to reduce unscheduled downtime and costly breakdowns. Empirical mode decomposition (EMD) is an adaptive time-frequency domain signal processing method, potentially suitable for nonstationary and/or nonlinear processes. However, the EMD method suffers from several problems such as mode mixing, defined as intrinsic mode functions (IMFs) with incorrect scales. In this paper, an ensemble noise-reconstructed EMD method is proposed to ameliorate the mode mixing problem and denoise IMFs for enhancing fault signatures. The proposed method defines the IMF components as an ensemble mean of EMD trials, where each trial is obtained by sifting signals that have been reconstructed using the estimated noise present in the measured signal. Unlike traditional denoising methods, the noise inherent in the input data is reconstructed and used to reduce the background noise. Furthermore, the reconstructed noise helps to project different scales of the signal onto their corresponding IMFs, instrumental in alleviating the mode mixing problem. Two critical issues concerned in the method, i.e., the noise estimation strategy and the number of EMD trials required for denoising are discussed. Furthermore, a comprehensive noise-assisted EMD method is proposed, which includes the proposed method and ensemble EMD (EEMD). Numerical simulations and experimental case studies on accelerometer data collected from an industrial shaving process are used to demonstrate and validate the proposed method. Results show that the proposed method can both detect impending faults and isolate multiple faults. Hence, the proposed method can act as a promising tool for mechanical fault detection.

In order to extract the vibration feature of remanufactured engines, the vibration signal is decomposed by ensemble empirical mode decomposition(EEMD), and this method is applied to study vibration mode of remanufactured engines. Based on the decomposition of vibration signal, correlation coefficient is introduced to study the correlation between IMF(Intrinsic Mode Function) components and original signal, and sensitive factors of IMF components are calculated using correlation coefficients. Hilbert transform is made with sensitive IMF components. The study results show that vibration feature of remanufactured engines is include in IMF components decomposed by EEMD, the vibration modes of remanufactured engines can be differentiated by Hilbert-Huang spectrum and marginal spectrum based on sensitive IMF components. According to vibration feature, the vibration state of remanufactured engines is divided into cylinder wall, cylinder cover and crankshaft vibration modes, which are important to improve the remanufacture level of engines.

Forecasting product sales in a consumer product company is one of the most important and challenging tasks for business financial planning and inventory management. In this paper, we are proposing a novel forecasting scheme through unique data pre-processing pipeline and machine learning model using a combination of Ensemble Empirical Mode Decomposition (EEMD) and Extreme Learning Machine (ELM) for multivariate sales data comprising of macroeconomic indicators. EEMD decomposition is performed on sales time series and intrinsic mode functions are analyzed. EEMD is also employed to find important constituent parts of indicators as a data pre-processing step. Experimental results show that EEMD helps in improving forecasting performance. We are also using auto-regressive sale terms along with macroeconomic indicators and employing feature ranking to analyze the most impactful features for prediction performance. Final forecasting performance is further compared with state of art techniques used for multivariate sales prediction models namely Vector Auto-Regressive (VAR) and Least Absolute Shrinkage and Selection Operator (LASSO) models. The proposed forecasting scheme demonstrates better performance than the current state of the art techniques.

In this paper a modified ensemble empirical mode decomposition (EEMD) method is presented, which is named winning-EEMD (W-EEMD). Two aspects of the EEMD, the amplitude of added white noise and the number of intrinsic mode functions (IMFs), are discussed in this method. The signal-to-noise ratio (SNR) is used to measure the amplitude of added noise and the winning number of IMFs (which results most frequency) is used to unify the number of IMFs. By this method, the calculation speed of decomposition is improved, and the relative error between original data and sum of decompositions is reduced. In addition, the feasibility and effectiveness of this method are proved by the example of the oceanic internal solitary wave.

Complex nonstationary and nonlinear signals can be adaptively analyzed by the Hilbert-Huang transform(HHT) through empirical mode decomposition(EMD) and the Hilbert transform to generate the Hilbert-Huang spectrum.The Hilbert-Huang spectrum is able to show the local characteristics of the signals and focus time-frequency information well,and therefore it is widely applied to analyzing vibration signals in the field of machinery fault diagnosis.EMD,however,sometimes cannot reveal the signal characteristics accurately because of the problem of mode confusion.Moreover,the extracted intrinsic mode functions(IMFs) have different sensitivity to faults.Some IMFs are sensitive and closely related to the faults but others are irrelevant.Thus it is necessary to enhance the fault diagnosis accuracy of the HHT based on all IMFs of EMD.Aiming at this problem,an improved HHT based on ensemble empirical mode decomposition(EEMD) and sensitive IMFs is proposed,in which EEMD is used to alleviate the problem of mode confusion,and the sensitivity evaluation algorithm to select sensitive IMFs and highlight potential fault characteristics.Thus,the fault diagnosis results obtained by the proposed method are more accurate than the HHT based on all IMFs of EMD.Both simulated signals and real signals of an early rub-impact fault verify the effectiveness of the improved HHT.

Background: Single-channel observed signal analysis based on independent component analysis (ICA) model belongs to the extremely underdetermined blind source separation (BBS) problem. Method: In order to extract the fault feature hidden in the single-channel measured signal from multi-stage gearbox, a joint approach of fault feature extraction based on ensemble empirical mode decomposition (EEMD) and constrained independent component analysis (CICA) is proposed. The single-channel vibration fault signal is decomposed into several intrinsic mode functions (IMFs) by EEMD, which can overcome the shortcomings of classical empirical mode decomposition (EMD). By computing the kurtosis and correlation coefficients of each IMF, we can select some suitable IMFs to construct a newly observed vector combined with the original signal, which meets the requirement of CICA algorithm. Finally, the suitable reference signal including faulty gear meshing frequency is generated, and the desired gear low frequency slight feature is extracted by CICA combined with envelope analysis. Conclusion: Through the experiment analysis of fault feature extraction with a missing tooth on the low speed gear pairs, the effectiveness and applicability of the proposed method are verified. Keywords: Gearbox, single-channel measured signal, ensemble empirical mode decomposition (EEMD), constrained independent component analysis (CICA), fault feature extraction, FFT spectrum.

A Hilbert–Huang transform (HHT) is a time–frequency technique and has been widely applied to analyzing vibration signals in the field of fault diagnosis of rotating machinery. It analyzes the vibration signals using intrinsic mode functions (IMFs) extracted using empirical mode decomposition (EMD). However, EMD sometimes cannot reveal the signal characteristics accurately because of the problem of mode mixing. Ensemble empirical mode decomposition (EEMD) was developed recently to alleviate this problem. The IMFs generated by EEMD have different sensitivity to faults. Some IMFs are sensitive and closely related to the faults but others are irrelevant. To enhance the accuracy of the HHT in fault diagnosis of rotating machinery, an improved HHT based on EEMD and sensitive IMFs is proposed in this paper. Simulated signals demonstrate the effectiveness of the improved HHT in diagnosing the faults of rotating machinery. Finally, the improved HHT is applied to diagnosing an early rub-impact fault of a heavy oil catalytic cracking machine set, and the application results prove that the improved HHT is superior to the HHT based on all IMFs of EMD.

Feature extraction is one of the most important, pivotal, and difficult problems in mechanical fault diagnosis, which directly relates to the accuracy of fault diagnosis and the reliability of early fault prediction. Therefore, a new fault feature extraction method, called the EDOMFE method based on integrating ensemble empirical mode decomposition (EEMD), mode selection, and multi-scale fuzzy entropy is proposed to accurately diagnose fault in this paper. The EEMD method is used to decompose the vibration signal into a series of intrinsic mode functions (IMFs) with a different physical significance. The correlation coefficient analysis method is used to calculate and determine three improved IMFs, which are close to the original signal. The multi-scale fuzzy entropy with the ability of effective distinguishing the complexity of different signals is used to calculate the entropy values of the selected three IMFs in order to form a feature vector with the complexity measure, which is regarded as the inputs of the support vector machine (SVM) model for training and constructing a SVM classifier (EOMSMFD based on EDOMFE and SVM) for fulfilling fault pattern recognition. Finally, the effectiveness of the proposed method is validated by real bearing vibration signals of the motor with different loads and fault severities. The experiment results show that the proposed EDOMFE method can effectively extract fault features from the vibration signal and that the proposed EOMSMFD method can accurately diagnose the fault types and fault severities for the inner race fault, the outer race fault, and rolling element fault of the motor bearing. Therefore, the proposed method provides a new fault diagnosis technology for rotating machinery.