Denoising of chaotic signal using independent component analysis and empirical mode decomposition with circulate translating

In this paper, a new method to reduce noises within chaotic signals based on ICA （independent component analysis） and EMD （empirical mode decomposition） is proposed. The basic idea is decomposing chaotic signals and constructing multidimensional input vectors, firstly, on the base of EMD and its translation invariance. Secondly, it makes the indepen- dent component analysis on the input vectors, which means that a self adapting denoising is carried out for the intrinsic mode functions （IMFs） of chaotic signals. Finally, all IMFs compose the new denoised chaotic signal. Experiments on the Lorenz chaotic signal composed of different Gaussian noises and the monthly observed chaotic sequence on sunspots were put into practice. The results proved that the method proposed in this paper is effective in denoising of chaotic signals. Moreover, it can correct the center point in the phase space effectively, which makes it approach the real track of the chaotic attractor.

Rolling element bearing instantaneous rotational frequency estimation based on EMD soft-thresholding denoising and instantaneous fault characteristic frequency

The accurate estimation of the rolling element bearing instantaneous rotational frequency(IRF) is the key capability of the order tracking method based on time-frequency analysis. The rolling element bearing IRF can be accurately estimated according to the instantaneous fault characteristic frequency(IFCF). However, in an environment with a low signal-to-noise ratio(SNR), e.g., an incipient fault or function at a low speed, the signal contains strong background noise that seriously affects the effectiveness of the aforementioned method. An algorithm of signal preprocessing based on empirical mode decomposition(EMD) and wavelet shrinkage was proposed in this work. Compared with EMD denoising by the cross-correlation coefficient and kurtosis(CCK) criterion, the method of EMD soft-thresholding(ST) denoising can ensure the integrity of the signal, improve the SNR, and highlight fault features. The effectiveness of the algorithm for rolling element bearing IRF estimation by EMD ST denoising and the IFCF was validated by both simulated and experimental bearing vibration signals at a low SNR.

Noise level estimation method with application to EMD-based signal denoising

This paper proposes a new signal noise level estimation approach by local regions. The estimated noise variance is applied as the threshold for an improved empirical mode decomposition（EMD） based signal denoising method. The proposed estimation method can effectively extract the candidate regions for the noise level estimation by measuring the correlation coefficient between noisy signal and a Gaussian filtered signal. For the improved EMD based method, the situation of decomposed intrinsic mode function（IMFs） which contains noise and signal simultaneously are taken into account. Experimental results from two simulated signals and an X-ray pulsar signal demonstrate that the proposed method can achieve better performance than the conventional EMD and wavelet transform（WT） based denoising methods.

An algorithm for ^(252)Cf-Source-Driven neutron signal denoising based on Compressive Sensing

As photoelectrically detected ^(252)Cf-source-driven neutron signals always contain noise, a denoising algorithm is proposed based on compressive sensing for the noised neutron signal. In the algorithm, Empirical Mode Decomposition(EMD) is applied to decompose the noised neutron signal and then find out the noised Intrinsic Mode Function(IMF) automatically. Thus, we only need to use the basis pursuit denoising(BPDN) algorithm to denoise these IMFs. For this reason, the proposed algorithm can be called EMDCSDN(Empirical Mode Decomposition Compressive Sensing Denoising). In addition, five indicators are employed to evaluate the denoising effect. The results show that the EMDCSDN algorithm is more effective than the other denoising algorithms including BPDN. This study provides a new approach for signal denoising at the front-end.

De-noising of radiation pressure signal generated by bubble oscillation based on ensemble empirical mode decomposition

The radiation pressure signals generated by the bubble oscillation are often utilized to recognize the characteristics of the target objects in many fields.However,these signals are easily contaminated by complex background noises.In order to accurately extract the effective components of the radiation pressure signal generated by the bubble oscillation,this paper proposes a de-noising procedure for the radiation pressure signal,based on the ensemble empirical mode decomposition(EEMD),the autocorrelation function and the modified wavelet soft-threshold de-noising method.In order to verify the effectiveness of the procedure,the typical radiation pressure signal generated based on the Keller-Miksis model under the acoustic excitation is employed for the subsequent de-noising analysis.The results of the qualitative analysis show that the amplitude and the period of the bubble oscillation can be clearly observed in the time-domain diagram of the de-noised signal based on the EEMD.In the quantitative analysis,the de-noised signal based on the EEMD has better performance with higher signal-to-noise ratio(SNR),smaller root-mean-square error,and larger correlation coefficient than that based on the wavelet transform(WT)and the empirical mode decomposition(EMD).Furthermore,with the increase of the complexity of the radiation pressure signal(e.g.,the increase of the dimensionless pressure amplitude of the acoustic wave and the decrease of the SNR of the input signal),the above three evaluation indexes of the de-noised signal based on the EEMD are all better than those based on the other two methods.When the signal is more complex,the de-noising capabilities of the WT,the EMD are greatly reduced,but the EEMD can still maintain the good de-noising capability,which shows the superiority of the signal de-noising procedure proposed in the present paper.

基于MI-EMD的激光引信回波信号去噪方法研究

为解决激光引信回波信号容易被噪声污染、信噪比低的问题,提出一种基于互信息经验模态分解的激光引信回波信号去噪方法。该方法融合了互信息相关性和经验模态分解自适应性的特点,对基于雷达原理建立的激光引信回波信号进行经验模态分解,通过互信息及相关阈值区分噪声模态和信号模态,提取出分解信号中的有用信号分量,并将其相关模态进行重构实现噪声的有效去除。实验结果表明:该方法处理后的信噪比提高到了18.8854 dB,均方根误差减小到2.81×10^(-6),且去噪后信号曲线的平滑度较高。该方法能有效滤除激光引信回波信号中的噪声,很好地还原激光引信回波原始信号,保证信号的完整性,为后续激光引信在噪声条件下的精确定距奠定了坚实的基础。

基于变分模态分解的肺音去噪算法

目的:为有效提高肺音信号质量,提出一种基于变分模态分解的肺音去噪方法。方法:首先利用经验模态分解对带噪肺音信号进行分解,根据本征模态函数特征确定最佳分解层数,然后根据分解层数对原始带噪肺音进行变分模态分解处理,接着根据皮尔逊系数选取有用模态,最后采用阈值方法对各模态函数去噪,重构后得到没有噪声干扰的肺音信号。结果:通过与维纳滤波和FIR滤波进行对比,本文方法的语音质量感知评价、短时间客观可读性和源信号失真比均更优。结论:本文方法能有效对肺音信号进行去噪处理。

基于WPD-EMD-WPD的地下工程微震信号降噪方法研究

微震信号中的背景噪声会影响初至拾取、震源定位及参数反演,合理有效的降噪方法是微震监测技术成功应用于工程建设的重要基础。本文提出一种基于WPD-EMD-WPD的方法抑制噪声,并采用信偏比衡量降噪效果。该方法首先对含噪信号小波包预降噪,实现初次滤波;然后对预降噪后的信号进行经验模态分解(Empirical Mode Decomposition,EMD),自适应分解得到IMFS,通过相关系数法确定IMFS分解分量位置;最后,对分界分量之前的高频分量小波包降噪,再与低频分量重构。分别使用小波包、EMD、EMD-WPD、WPD-EMD-WPD 4种方法进行仿真实验,对含噪Ricker子波降噪处理,通过对比降噪前后的降噪效果衡量指标、频谱图、波形图对比发现,WPD-EMD-WPD降噪效果更优,且信偏比与其他降噪指标有良好对应性。将该方法应用于国内某大型水电工程地下洞室,结果表明,该方法能获得更低的信偏比,且能更好地反映初至时刻和岩石微破裂信息。

EMD端点效应问题解决方法的研究

经验模态分解(EMD)是近年来兴起的一种非平稳信号处理方法,但在分解过程中因产生端点效应而造成分解结果严重失真。因此,研究了一种自适应波形匹配延拓法的改进方法,该方法根据信号端部不同特点确定延拓点并进行延拓,从而解决端点效应问题。实验结果表明,改进的自适应波形匹配延拓法能避免信号EMD分解失真。将改进的自适应波形匹配延拓法的EMD结合小波软阈值去噪方法用于处理加噪超声仿真信号,得到的信噪比比小波软阈值法处理后的信噪比高33.78%,说明该方法在非线性、非平稳信号处理方面有一定优势。

基于集合经验模态分解的心电信号自适应降噪及基线漂移修正

在心电信号的采集过程中,各种噪声的干扰会引起信号失真及基线漂移,进而影响对心脏信号的精准判断。针对此,提出一种基于集合经验模态分解的自适应算法。首先,对含有噪声及基线漂移的心电信号进行集合经验模态分解(EEMD),分解出固有模态函数(IMF)分量。然后,筛选出需要处理的IMF分量。最后,通过自适应窗口处理带噪的低阶IMF以及移除导致基线漂移的高阶IMF,从而达到降噪和修正基线漂移的目的。在MIT-BIH数据库中的实验结果表明,基于EEMD方法的降噪效果良好,在同等肌电噪声情况下,与基于EMD的自适应窗口法对比,在平均信噪比上提升1.7507,增幅约为13%;在同等基线漂移情况下,与基于EEMD的阈值法对比,在平均基线矫正率上下降0.0795,降幅约为14%。

基于EEMD、相关系数、排列熵和小波阈值去噪的新型水下声学信号去噪方法

The complexities of the marine environment and the unique characteristics of underwater channels pose challenges in obtaining reliable signals underwater,necessitating the filtration of underwater acoustic noise.Herein,an underwater acoustic signal denoising method based on ensemble empirical mode decomposition(EEMD),correlation coefficient(CC),permutation entropy(PE),and wavelet threshold denoising(WTD)is proposed.Furthermore,simulation experiments are conducted using simulated and real underwater acoustic data.The experimental results reveal that the proposed denoising method outperforms other previous methods in terms of signal-to-noise ratio,root mean square error,and CC.The proposed method eliminates noise and retains valuable information in the signal.

CEEMD结合改进小波阈值的激光雷达信号去噪算法

激光雷达远距离回波信号受噪声影响,严重失真。为了有效去除信号的噪声,提高回波信号信噪比,提出一种互补集合经验模态分解(complementary ensemble empirical mode decomposition,CEEMD)结合改进小波阈值的去噪算法。CEEMD可以自适应地分解非线性和非平稳信号,改进小波阈值函数具有高阶可导特性,能够克服硬阈值、软阈值函数各自存在的问题。两种方法结合,可以更有效地去除噪声。首先,对回波信号进行CEEMD分解,得到若干固有模态函数(intrinsic mode function,IMF)。其次,通过相关系数法计算IMF分量与信号的相关系数,确定相关分量和不相关分量。最后,对不相关分量使用小波改进阈值法进行去噪,对相关分量使用粗糙惩罚法进行平滑,再重构信号。基于实测数据的实验结果表明,所提算法比CEEMD去噪法和CEEMD结合原改进阈值去噪法,信噪比分别提升了2.65 dB和0.58 dB。

基于改进核极限学习机的风电功率短期预测

针对环境变化造成风力发电功率波动大和核极限学习机易陷入局部最优解的问题,构建了一种基于完全噪声辅助聚合经验模态分解(complete ensemble empirical mode decomposition with adaptive noise analysis,CEEMDAN)、小波阈值去噪和粒子群算法优化核极限学习机的风电功率短期预测模型。首先,利用CEEMDAN对风力发电输出功率密切相关的环境因素进行分解,得到若干个规律性较强的模态分量,利用阈值去噪法对含噪声较多的第一模态分量进行去噪,削弱环境因素的非平稳性;然后,将分解后的子分量和风电功率历史数据作为粒子群优化后的核极限学习机算法的输入进行预测;最后,选用河北张家口某风电场的实测数据进行实验对比分析。实验结果表明:所提出的改进风电功率预测组合模型的预测精度更高,适应于不同季节环境下的风电功率预测。

基于EMD-PSO优化阈值的COD光谱去噪研究

在对水体COD进行光谱检测时,光谱数据中会掺杂大量噪声,影响后续对水体COD的定量分析。为了降低噪声的影响,提出了一种经验模态分解-粒子群(EMD-PSO)优化阈值去噪模型,首先对EMD分解得到的特征模函数进行噪声主导分量和信号主导分量的划分,通过PSO优化选取噪声主导分量部分的最佳阈值,对数据进行截取、重构,获得去噪光谱信号。以不同浓度梯度的COD标准溶液进行测试,实验结果表明:EMDPSO优化阈值去噪模型相对于EMD软、硬阈值去噪,信噪比有所提高,均方误差由软、硬阈值的0.6247和0.6725降至0.02619,样品浓度与吸光度线性相关系数更高,达到0.99798。

基于CEEMDAN-SE-TCN的集群资源预测研究

针对服务器集群负载数据的波动性和非线性特点,提出一种基于CEEMDAN-SE-TCN的预测算法。该算法首先将原始服务器集群数据经过自适应加噪集合经验模态分解(CEEMDAN),有效降低负载序列复杂度。然后,在得到分解后的相关IMF分量后,利用相关系数法将各IMF分量与原始序列进行比较,去除相关性较弱的分量。最后,提取各分量相应的特征值输入并加入注意力机制的时间卷积网络(SE-TCN)进行建模预测。通过Google集群数据集中的CPU负载率序列实测证明,在同等条件下CEEMDAN-SE-TCN模型整体优于其他基准模型,MAPE指标相较于其他模型分别降低7.1%、6.5%、2.5%,证明了该算法的有效性和可行性。

基于二次CEEMD与时域特征分析的去噪方法

为克服经验模态分解(EMD)去噪方法存在的模态混叠以及噪声分量与信号分量区分困难问题,本文提出了一种基于二次互补集合经验模态分解(CEEMD)与时域特征分析的去噪方法。该方法利用CEEMD来克服模态混叠问题,同时基于对CEEMD本征模态函数(IMF)的时域特征分析来确定噪声主导IMF分量与信号主导IMF分量的分界点,据此区分噪声分量与信号分量,并对分界点相邻两侧的噪声主导IMF分量与信号主导IMF分量进行二次CEEMD分解,在保留更多有用信号的同时进一步滤除剩余噪声。对含冲击噪声干扰的实际机载平台数据的去噪实验结果表明,新方法通过对噪声分量与信号分量的有效分离,可以更好地抑制噪声干扰,明显提升信噪比。

基于CEEMDAN-小波包自适应阈值混凝土声发射信号降噪研究

为了得到更加纯净的混凝土声发射(acoustic emission, AE)信号来更准确地监测混凝土结构破裂过程,提出了一种完全自适应噪声集合经验模态分解(complete ensemble empirical mode decomposition, CEEMDAN)与小波包自适应阈值联合方法对循环荷载作用下的混凝土声发射信号进行降噪处理,运用信噪比和快速傅里叶变化(fast Fourier transform, FFT)分析来验证所用方法的可行性。实验结果表明:结合CEEMDAN-小波包自适应阈值对混凝土声发射信号进行降噪的效果较好,能有效地保留混凝土声发射信号特征信息,对混凝土声发射信号降噪提供新的思路,为后续利用声发射信号分析混凝土结构内部微裂纹扩展及演化特征奠定基础。

基于EMD-NLMS的船舶发动机声信号降噪处理方法

针对船舶声信号降噪滤波的问题,提出了一种基于经验模态分解-归一最小均方的算法。该方法进行经验模态分解得出噪声分量,将得到的噪声分量作为输入信号进行自适应滤波,通过自适应滤波算法迭代处理得到降噪后的信号分量,并把信号分量叠加得到最终降噪后的信号。通过对比最小均方算法、归一化最小均方算法、经验模态分解-最小均方算法和经验模态分解-归一最小均方算法对船舶声信号的降噪效果,得出在船舶声信号滤波降噪方面经验模态分解-归一最小均方算法相比于其他三种算法有更好的滤波效果。

CEEMDAN联合小波阈值算法在水下激光雷达中抑制散射杂波的应用

水下激光雷达回波信号中往往含有大量散射噪声.为了能够有效抑制散射噪声,提高水下激光雷达测距精度,提出了基于自适应完备噪声经验模态分解(CEEMDAN)与小波阈值相结合的去噪新方法.首先通过相关系数法对自适应完备噪声经验模态分解得到的本征模态函数(IMF)进行筛选;然后对筛选后的本征模态函数进行小波阈值去噪,进一步去除本征模态函数中的噪声成分;最后将去噪后的本征模态函数进行信号重构得到去噪后信号.将该方法应用到不同衰减系数水体的强度调制连续光水下测距实验,使用白色聚氯乙烯(PVC)反射板为探测目标,在3.75个衰减长度时,直接采用相关极值确定延时,测距误差达到19.2 cm;应用该方法处理后,测距误差减小到6.2 cm,有效提高测距精度.

基于经验模态分解的汽车轮速信号去噪算法

为了提升车辆行驶的安全性,保障车辆的稳健性和判断力,提出一种基于经验模态分解的汽车轮速信号去噪算法。分析信号内部能量和频率变化,为改善时频性能将数据拆解为窄带信号。通过最小尺度筛分去除噪声低阶分量后,再对剩下分量的进行重构。受影响严重时信号与噪声会呈现混叠模态,提取正常信号本征模态函数分量,通过经验模态分解找到信号索引值,以此过滤汽车轮速含噪IMF分量,完成信号去噪。通过仿真,证明所提算法能够有效将信号去噪,去除误差最小,效果最优。