Research on EMD End Effect Based on LS-SVM Regression

王亚东,郝志勇

DOI: https://doi.org/10.3969/j.issn.1674-7046.2012.04.015

2012-01-01

Abstract:In this paper, regarding the problem of EMD end effect, using the least square support vector re- gression, it could predict the signal＇s ends to inhibit the end effect, the Hilbert spectrum after Endpoint ex- tension is discussed based on the analysis of simulation signal.

What problem does this paper attempt to address?

Application of improved LS-SVM method in end effect of EMD

Zhijun XU,Huan KUANG,Rulong WANG,Baojian HUA

DOI: https://doi.org/10.3778/j.issn.1002-8331.1305-0500

2015-01-01

Abstract:The empirical mode decomposition is an effective method for non-stationary and nonlinear signals processing, but at the same time it produces endpoint effects which would decrease the decomposition accuracy when the three order spline interpolation is used repeatedly. In order to suppress the endpoint effect in empirical mode decomposition, this paper proposes an approach based on least squares support vector machine and mirror extension. In this new method, by using LS-SVM to the original signal sequence it would produce a finite number of extensional data points at both ends, and then with mirror extension acting on the extensional data a symmetrical continuation signal sequence would be formed. Then a ring signal sequence would be produced after the end extension and mirror extension. The empirical mode decomposition is used to deal with the ring signal sequence. By analyzing the simulation signals and real EEG signals with the new method, the results show that the method can effectively restrain the endpoint effect. And compared with other extension methods, it is better than Support Vector Machine(SVM), and the Least Squares Support Vector Machine(LS-SVM).
A New Method for Restraining the End Effect of Empirical Mode Decomposition and Its Applications to Signal Feature Extraction

曹冲锋,杨世锡,杨将新

DOI: https://doi.org/10.3969/j.issn.1004-4523.2008.06.009

2008-01-01

Abstract:The end effect of empirical mode decomposition(EMD) was always a difficult and significant problem,based on the consideration of the mechanism of the end effect and the current research development,we presented a new method of End optimization symmetric extension,through which the end effect is restrained.Firstly,the assumption is that the two ends of original signal are unknown,and which is extended in a point-symmetrical manner with end-point as its center.Secondly,a deviation error evaluation function concerning original signal is constituted and minimized,then the two optimization end-points value are obtained.Extended in a point symmetrical manner with new end-point value,the obtained up and down envelops maximally approached original signal end-point,which restrained the envelops divergence in EMD algorithm.Finally,in the process of filtering intrinsic mode function(IMF) discard the extended data,which release end effect to original signal outside and maximally reduced distortion of original signal.Compared the three methods named EMD,EMD based on EOSE and EMD based on ARMA,simulation and experiment show that the proposed method could restrain end effect effectively and precisely extract classic fault feature of vibration signal of rotating machine.
Evaluation of the EMD End Effect and a Window Based Method to Improve EMD

Ren, Daqian,Yang, Shixi,Wu, Zhaotong,Yan, Gongbiao

DOI: https://doi.org/10.1049/cp:20061016

2006-01-01

Abstract:The end effect is a key shortcoming of the Empirical Mode Decomposition(EMD). On the sifting process, the upper and lower envelop can have large swing. The end effect leads to fake components and an over-estimated energy of the signal using EMD.A relative deviation of the true signal energy and the IMFs energy after EMD has been developed to evaluate the end effect. The end effect can be reduced by applying a window function to the signal. The experiment result shows the method is feasible.
Research on End Effect of LMD Based Time-frequency Analysis in Rotating Machinery Fault Diagnosis

任达千,杨世锡,吴昭同,严拱标

DOI: https://doi.org/10.3969/j.issn.1004-132X.2012.08.015

2012-01-01

Abstract:An index was developed based on energy for evaluating the end effect of LMD.The end effects of LMD and EMD（empirical mode decomposition） were compared.Because the envelop definition was different,in LMD the undefined envelop near border was shorter than that in EMD.LMD end effect is not as seriously as EMD.A mirror extension method was developed to reduce the end effect.The simulation results are quite good.LMD is applied in cracked rotor diagnosis,and the experimental results are also quite good.
Processing method of EMD endpoint effect based on SVRM extension

Kai Chai,Shao-Wei Feng

DOI: https://doi.org/10.21595/vp.2020.21772

2020-11-26

Vibroengineering PROCEDIA

Abstract:Aiming at the problem of endpoint effect in empirical mode decomposition (EMD), the application method of support vector regression machine (SVRM) in EMD extension data prediction is studied. Firstly, the basic principle, data extension method and parameter setting of SVRM are introduced. Secondly, several application methods of SVRM in EMD extension are studied to analyze and verify the operational efficiency and decomposition accuracy characteristics of each method respectively. Finally, the proposed extension method based on SVRM extreme value point prediction can greatly improve the operation efficiency of SVRM long time extension. The simulation signal analysis shows that the SVRM extreme point prediction extension method not only improves the accuracy and reliability of EMD decomposition, but also effectively inhibits the end-point effect phenomenon, significantly reduces the SVRM extension time, and improves the practicability of EMD method.
Boundary-extension Method in Hilbert-Huang Transform

沈路,周晓军,张志刚,张文斌

DOI: https://doi.org/10.3969/j.issn.1000-3835.2009.08.038

2009-01-01

Abstract:Serious end effect will be produced when the upper and lower envelopes are constructed in the process of EMD. Similar extrema extension method was proposed to overcome the problem. Abscissa difference of adjacent extrema at both ends of data was considered as the shape characteristic, then the shape characteristic and the average of extrema were used to extend signal. Both the trend of extrema variation near boundary and the rest of extrema were considered in similar extrema extension, while only the extrema near boundary are taken to extend signal in current methods. The simulation and experiment results show the method is effective in restraining end effect, furthermore it is simple and easy to realize.
Improved Weighted Least Squares Support Vector Machines Algorithm And Its Applications In Spectroscopic Quantitative Analysis

Jian Feng Lü,Dai Lian-Kui

DOI: https://doi.org/10.3321/j.issn:0253-3820.2007.03.006

IF: 1.193

2007-01-01

Chinese Journal of Analytical Chemistry

Abstract:An improved algorithm of weighted least squares support vector machines (WLS-SVM) is proposed to overcome the negatine influence of abnornmal traningsamples on calibration models. The improved algorithm solves the iterative convergence problem in the original algorithm and has been used in spectroscopic quantitative analysis. Compared to the original algorithm, the experimental results show that the new algorithm remarkably improves model robustness and the ability to detect abnormal samples.
Robust Iterative Algorithm of Weighted Least Squares Support Vector Machine and Its Application in Spectral Analysis

Bao Xin,Dai Liankui

DOI: https://doi.org/10.3321/j.issn:0567-7351.2009.10.008

2009-01-01

Acta Chimica Sinica

Abstract:A robust iterative algorithm of weighted least squares support vector machine (WLS-SVM) was proposed to overcome the negative influence of outliers on spectral analysis. A novel method to calculate regression error was proposed to solve the iterative convergence problem in original WLS-SVM; to improve the robustness of original WLS-SVM, the formula for computing weighted value was also revised: the median value of regression error was selected as criteria. This algorithm has been applied to spectral quantitative analysis. Compared with original algorithm, the experimental results show the proposed algorithm is convergent and the breakdown point value is approximately 35%.
M-estimator Based Support Vector Machine and Its Application

BAO Xin,DAI Liankui

DOI: https://doi.org/10.3321/j.issn:0438-1157.2009.07.021

2009-01-01

Abstract:Validity of training samples is important to regression model,however,outliers in training sample sets are hard to avoid,which will disturb the model accuracy.In order to overcome the influence of outliers to regression model,a new M-estimator based support vector machine(M-SVM) was introduced.It replaced the sum of square residuals in least squares support vector machine(LS-SVM) with objective function of M-estimator.An iterative modeling algorithm of the M-SVM was proposed.The M-SVM was used in a low-dimension data regression problem with outliers and in the NIR spectral analysis of gasoline research octane number.Experimental results showed that the M-SVM was more robust and accurate than other SVMs.
Iterative robust least square support vector machine for spectral analysis

Xin Bao,Dai Liankui

2010-01-01

Asian Journal of Chemistry

Abstract:The aim of this study is to develop a novel robust regression algorithm: robust least squares support vector machine (RLS-SVM), to overcome the limitation of the existing support vector machine at high percent of contamination for spectral analysis. In the algorithm, firstly a subset is selected randomly from the original data set to build regression model and the robust estimates of the residuals for the whole set are generated; then the confidence interval of the residuals distribution is applied iteratively to detect outliers. Finally, the LS-SVM estimates are created from the regression model being trained with the selected subset without outliers. The proposed algorithm is applied in the near infrared spectral analysis of gasoline samples in order to predict their octane number with some outliers. Compared with other support vector machine algorithms, the test results show the breakdown point value for the algorithm can be over 45 %. The results also show its priority in predicted precision.
Semi-Quantitative Modeling for EMD

CHENXI SHAO,ZHIZHI ZHU,QIYU SHAO,XINGFU WANG,FUYOU MIAO,WEIHUA WANG,YING ZHANG

DOI: https://doi.org/10.12783/dtcse/iceiti2017/18898

2018-01-01

DEStech Transactions on Computer Science and Engineering

Abstract:In this paper, fully considering that the end effect is an uncertain problem in essence, we proposed a semi-quantitative modeling method, called SQME, to mitigate the effect. By SQME, the qualitative behaviors and the extreme points at both ends are predicted well in the time domain, and the data extension is more reasonable. The feasibility of SQME is illustrated in theory. The simulation and experiment results show that SQME can mitigate the end effect of EMD effectively, and the standard error between the extended series and the actual series is controlled within 1%.
Suppressing End Effect of EMD Based on Local Polynomial Regression

LI Yong,FANG Zhaoben,WEI Yongfeng

DOI: https://doi.org/10.3969/j.issn.0253-2778.2014.09.013

2014-01-01

Journal of University of Science and Technology of China

Abstract:Hilbert-Huang transform (HHT ) is a completely data-driven adaptive algorithm for analyzing time-frequency data .HHT has been widely used in many areas .However ,as a core part of the HHT , empirical mode decomposition (EMD ) is still negatively affected by end effect . Based on the local polynomial regression (LPR) ,one extreme point was added to each of extreme series by few local data to suppress the end effect of EMD .The simulation results show that the LPR-EMD is a practical algorithm that can effectively restrain the end effect and have less constrain on the data and run faster .
End Effects Mitigation For Empirical Mode Decomposition With Nonlinear Gray Model

Zhi He,Yi Shen,Qiang Wang,Yan Wang

DOI: https://doi.org/10.1142/S1793536912500021

2012-01-01

Advances in Data Science and Adaptive Analysis

Abstract:To mitigate end effects of empirical mode decomposition (EMD), a novel approach inspired by the nonlinear gray model (GM) termed as GM(1,1, alpha) is presented. Other than traditional linear or mirror extension on the boundary, the GM(1,1, alpha) model is applied to predict two extrema at both ends of the data. It is worth noting that our GM(1,1, alpha) model is particularly useful for predicting uncertainty data. According to numerical experiments on synthetic signal as well as real data series, the proposed method gives very comparable results with other three generally acknowledged methods, including the linear extension (LE), window function (WF), and mirror symmetry (MS) based methods. That is, the proposed method can reduce end effects and improve decomposition results of EMD significantly.
Performance Analysis of Least Squares Support Vector Regression Filtering System

Deng Xiao-Ying,Yang Ding-Hui,Liu Tao,Li Yue,Yang Bao-Jun

DOI: https://doi.org/10.3969/j.issn.0001-5733.2010.08.027

2010-01-01

Abstract:Support vector machine (SVM) is always researched and developed as a machine learning method on the base of statistical learning theory. As viewed from signal and system, the least squares support vector machine (LS-SVM) with the translation invariant kernel is a linear time invariant system. Taking the Ricker wavelet kernel as an example, we investigate the effects of different parameters on frequency responses of the least squares support vector regression (LS-SVR) filter. Those parameters affect the rising edge, the band width and central frequency of passband, and also the attenuation of signal energy. In other words, the longer the length of LS-SVR filter, the sharper the rising edge generated; the larger the kernel parameter, the higher the central frequency and the wider the bandwidth of the passband; the smaller the regularization parameter, the narrower the bandwidth of passband and the greater the attenuation of the desired signal. The experimental results of synthetic seismic data show that the LS-SVR filter with the Ricker wavelet kernel works better than the LS-SVR filter with the RBF kernel, the wavelet transform-based method and adaptive Wiener filtering method.
Detection Method for Walnut Shell-Kernel Separation Accuracy Based on Near-Infrared Spectroscopy.

Minhui An,Chengmao Cao,Zhengmin Wu,Kun Luo

DOI: https://doi.org/10.3390/s22218301

IF: 3.9

2022-01-01

Sensors

Abstract:In this study, Near-infrared (NIR) spectroscopy was adopted for the collection of 1200 spectra of three types of walnut materials after breaking the shells. A detection model of the walnut shell-kernel separation accuracy was established. The preprocessing method of de-trending (DT) was adopted. A classification model based on a support vector machine (SVM) and an extreme learning machine (ELM) was established with the principal component factor as the input variable. The effect of the penalty value (C) and kernel width (g) on the SVM model was discussed. The selection criteria of the number of hidden layer nodes (L) in the ELM model were studied, and a genetic algorithm (GA) was used to optimize the input layer weight (W) and the hidden layer threshold value (B) of the ELM. The results revealed that the classification accuracy of SVM and ELM models for the shell, kernel, and chimera was 97.78% and 97.11%. The proposed method can serve as a reference for the detection of walnut shell-kernel separation accuracy.
Active Machine Learning for Automatic High-Frequency EMI Source Localization

Jinghai Guo,Ling Zhang,Xin Yan,Hanzhi Ma,Da Li,Er-Ping Li

DOI: https://doi.org/10.1109/emcsipi49824.2024.10705449

2024-01-01

Abstract:This paper proposes a sparse emission source microscopy (ESM) technique based on active machine learning to localize high-frequency electromagnetic interference (EMI) sources. The proposed method, which combines the query-by-committee (QBC) algorithm and a newly proposed entropy feedback (EF) technique, can significantly improve the efficiency of high-frequency EMI source location by reducing the number of scanning points. The results of simulation and measurement examples show that the proposed method is much more time-efficient compared with the full sampling method. Also, the method demonstrates a noticeable advantage over the uniform sampling and random sampling approaches, and has higher accuracy of EMI source location with the same number of sparse scanning samples.
Research On Soft Sensor Based On Extreme Learning Machine

Zhili Xiong,Shaocheng Qu,Zhigao Lan

DOI: https://doi.org/10.1109/ccdc.2017.7978733

2017-01-01

Abstract:the traditional ELM algorithm in the hidden layer part parameters when the data sets are determined randomly and outliers are present, using the least squares method often exaggerate the singular value influence, cause the system deviation, resulting in the results of ELM network instability. In view of the above problems, an improved algorithm for extreme learning machine is proposed. By introducing the M estimation method, the specific idea is iterative weighted least squares estimation of regression coefficient, so as to optimize the residual sum of squares objective function to determine the weights of each sample to improve the robustness of traditional extreme learning machine. Finally, the data set is constructed to validate the proposed method, and the results show that the proposed method can reduce the impact of outliers and improve the prediction accuracy.
Elimination of End Effects in LMD Based on LSTM Network and Applications for Rolling Bearing Fault Feature Extraction

Jianhong Liang,Liping Wang,Jun Wu,Zhigui Liu

DOI: https://doi.org/10.1155/2020/7293454

IF: 1.43

2020-01-01

Mathematical Problems in Engineering

Abstract:Local mean decomposition (LMD) is widely used in the area of multicomponents signal processing and fault diagnosis. One of the major problems is end effects, which distort the decomposed waveform at each end of the analyzed signal and influence feature frequency. In order to solve this problem, this paper proposes a novel self-adaptive waveform point extended method based on long short-term memory (LSTM) network. First, based on existing signal points, the LSTM network parameters of right and left ends are trained; then, these parameters are used to extend the waveform point at each end-side of signal; furthermore, the corresponding parameters are adaptively updated. The proposed method is compared with the characteristic segment extension and the traditional neural network extension methods through a simulated signal to verify the effectiveness. By combing the proposed method with LMD, an improved LMD algorithm is obtained. Finally, application of rolling bearing fault signal is carried out by the improved LMD algorithm, and the results show that the feature frequencies of the rolling bearing's ball and inner and outer rings are successfully extracted.
AR Model Prediction-Based EMD Method and Its Application to Data Extension

Jingsong Hu,Shixi Yang

DOI: https://doi.org/10.3969/j.issn.1004-6801.2007.02.006

2007-01-01

Abstract:The AR model prediction-based data extension technology was introduced to the field of empirical mode decomposition (EMD) time-frequency analysis. The data extension technology is that, first modeling the original data with the AR method, then extending the data with the construct model. A simulate nonlinear signal is researched with the method and results show that the technology is valid in the field of the EMD time-frequency analysis. A deep crack rotor vibration signal is researched also using the extension method and the results show that the method is helpful to detect the phase-modulation caused by torsional vibration of the rotor. The result of the study can be widely used in time-frequency analysis field.
Elimination of End Effects in Local Mean Decomposition Using Spectral Coherence and Applications for Rotating Machinery.

Wei Guo,Lingjian Huang,Cheng Chen,Huaiwu Zou,Zhiwen Liu

DOI: https://doi.org/10.1016/j.dsp.2016.04.007

IF: 2.92

2016-01-01

Digital Signal Processing

Abstract:Local mean decomposition (LMD) is widely used in signal processing and fault diagnosis of rotating machinery as an adaptive signal processing method. It is developed from the popular empirical mode decomposition (EMD). Both of them have an open problem of end effects, which influences the performance of the signal decomposition and distort the results. Using the cyclostationary property of a vibration signal generated by rotating machinery, a novel signal waveform extension method is proposed to solve this problem. The method mainly includes three steps: waveform segmentation, spectral coherence comparison, and waveform extension. Its main idea is to automatically search the inside segment having similar frequency spectrum to one end of the analyzed signal, and then use its successive segment to extend the waveform, so that the extended signal can maintain temporal continuity in time domain and spectral coherence in frequency domain. A simulated signal is used to illustrate the proposed extension method and the comparison with the popular mirror extension and neural-network-based extension methods demonstrates its better performance on waveform extension. After that, combining the proposed extension method with normal LMD, the improved LMD method is applied to three experimental vibration signals collected from different rotating machines. The results demonstrate that the proposed waveform extension method based on spectral coherence can well extend the vibration signal, accordingly, errors caused by end effects would not distort the signal as well as its decomposition results.

Research on EMD End Effect Based on LS-SVM Regression

Application of improved LS-SVM method in end effect of EMD

A New Method for Restraining the End Effect of Empirical Mode Decomposition and Its Applications to Signal Feature Extraction

Evaluation of the EMD End Effect and a Window Based Method to Improve EMD

Research on End Effect of LMD Based Time-frequency Analysis in Rotating Machinery Fault Diagnosis

Processing method of EMD endpoint effect based on SVRM extension

Boundary-extension Method in Hilbert-Huang Transform

Improved Weighted Least Squares Support Vector Machines Algorithm And Its Applications In Spectroscopic Quantitative Analysis

Robust Iterative Algorithm of Weighted Least Squares Support Vector Machine and Its Application in Spectral Analysis

M-estimator Based Support Vector Machine and Its Application

Iterative robust least square support vector machine for spectral analysis

Semi-Quantitative Modeling for EMD

Suppressing End Effect of EMD Based on Local Polynomial Regression

End Effects Mitigation For Empirical Mode Decomposition With Nonlinear Gray Model

Performance Analysis of Least Squares Support Vector Regression Filtering System

Detection Method for Walnut Shell-Kernel Separation Accuracy Based on Near-Infrared Spectroscopy.

Active Machine Learning for Automatic High-Frequency EMI Source Localization

Research On Soft Sensor Based On Extreme Learning Machine

Elimination of End Effects in LMD Based on LSTM Network and Applications for Rolling Bearing Fault Feature Extraction

AR Model Prediction-Based EMD Method and Its Application to Data Extension

Elimination of End Effects in Local Mean Decomposition Using Spectral Coherence and Applications for Rotating Machinery.