A review of vibration and acoustic measurement methods for the detection of defects in rolling element bearings

A review of vibration and acoustic measurement methods for the detection of defects in rolling element bearings is presented in this paper. Detection of both localized and distributed categories of defect has been considered. An explanation for the vibration and noise generation in bearings is given. Vibration measurement in both time and frequency domains along with signal processing techniques such as the high-frequency resonance technique have been covered. Other acoustic measurement techniques such as sound pressure, sound intensity and acoustic emission have been reviewed. Recent trends in research on the detection of defects in bearings, such as the wavelet transform method and automated data processing, have also been included.     

EMD方法在消除桥梁振动信号局部强干扰中的应用

在进行桥梁的健康监测和状态评估时,外界环境的影响常常会在采集的振动信号中形成局部强干扰,导致分析结果的严重失真。为解决这一问题,文中基于经验模态分解(empirical mode decomposition,EMD),提出一种信号强干扰的消除方法。首先利用EMD把一个时间序列的信号分解成不同时间尺度的本征模函数(intnnsic mode function,IMF)和残余项,然后采用合适的带通滤波器对前几个IMF进行滤波,在存在强干扰的区段,用滤波后的数据代替滤波前的数据,并使后几个IMF在相应区段的幅值为零,最后将所有的IMF及趋势项重新进行叠加,即得到消除强干扰后的信号,将该信号再次进行EMD分解,可得到一系列新的IMF,它与未消除干扰时信号的分解结果有显著差别。通过对实测南京桥有对讲机干扰的应变信号进行分析,结果表明该方法可行、有效。     

Trend extraction based on separations of consecutive empirical mode decomposition components in Hilbert marginal spectrum

Extracting the underlying trends is an important tool for the analysis of signals. This paper presents a novel methodology for extracting the underlying trends of signals based on the separations of consecutive empirical mode decomposition (EMD) components in the Hilbert marginal spectrum. A signal is initially represented as a sum of intrinsic mode functions (IMFs) obtained via the EMD. The Hilbert marginal spectrum of each IMF is then calculated. The separations of two consecutive IMFs in the Hilbert marginal spectrum are estimated based on their correlation coefficients. The group of the last several IMFs in which the IMFs are close to each other in the Hilbert marginal spectrum will be used for the representation of the underlying trend of the signal. Extensive experimental results are presented to illustrate the rationale and the effectiveness of the proposed method.     

齿轮传动系统非线性频率调制规律仿真分析及实验研究

针对传统快速傅里叶变换(FFT)研究齿轮系统非线性非平稳振动存在虚假信号和假频的问题,提出了基于希尔伯特黄变换的本征模态函数(IMFs)傅里叶变换法。该方法通过经验模态分解(EMD)将原始信号分解为一系列不同时间特征尺度的IMFs,对能够反映原始信号物理意义的IMF进行傅里叶变换。以某型采煤机截割部齿轮箱为工程范例,通过振动实验得到齿轮系统非线性振动响应。分别利用传统FFT法和本征模态函数FFT法对实测信号进行非线性振动分析。研究结果表明:本征模态函数FFT法显著减少了多余且无意义的频率成分,能够更好地识别参与非线性频率调制的齿轮啮合特征频率,避免了传统FFT法产生的虚假频率干扰。该研究对分析齿轮转子系统非线性频率调制现象具有一定的参考价值。     

GEARBOX FAULT DIAGNOSIS BASED ON EMPIRICAL MODE DECOMPOSITION

Time synchronous averaging of vibration data is a fundament technique for gearbox diagnosis. Currently, this technique relies on hardware tachometer to give phase synchronous information. Empirical mode decomposition (HMD) is introduced to replace time synchronous averaging of gearbox vibration signal. With it, any complicated dataset can be decomposed into a finite and often small number of intrinsic mode functions (IMF). The key problem is how to assure that vibration signals deduced by gear defects could be sifted out by HMD. The characteristic vibration signals of gear defects are proved IMFs, which makes it possible to utilize EMD for the diagnosis of gearbox faults. The method is validated by data from recordings of the vibration of a single-stage spiral bevel gearbox with fatigue pitting. The results show EMD is powerful to extract characteristic information from noisy vibration signals.     

基于EEMD分解和奇异值差分谱理论的轴承故障诊断研究

提出一种基于总体平均经验模态分解(ensemble empirical mode decomposition,EEMD)和奇异值差分谱的轴承故障诊断方法。首先将非平稳的原始轴承振动信号通过EEMD方法分解成若干个平稳的本征模函数(intrinsic modefunction,IMF);由于背景噪声的影响,从各个IMF的频谱中难以准确地得到故障频率。对IMF分量构建Hankel矩阵,并进行奇异值分解,进一步找到奇异值差分谱,根据奇异值差分谱理论对某IMF分量进行消噪和重构,然后再求其频谱,便能准确地得到故障频率。实验结果表明,所提出的方法能有效地应用于轴承的故障诊断。     

基于Hilbert包络谱奇异值和IPSO-SVM的滚动轴承故障诊断研究

针对表征滚动轴承故障信号特征难提取及支持向量机结构参数依据经验选取,致使故障分类模型的精度、泛化能力差的问题,提出一种基于Hilbert包络谱奇异值和改进粒子群（Improvedparticleswarmoptimization,IPSO）优化支持向量机（Supportvectormachine,SVM）的滚动轴承状态辨识方法。首先,利用经验模态分解（Empiricalmodedecomposition,EMD）所采集的滚动轴承信号,并将所获相关程度较大的本征模式分量（Intrinsicmodefunction,IMF）进行Hilbert解调包络分析来获取包络矩阵,并在此基础上进行奇异值分解。其次,利用IPSO算法优化SVM的惩罚系数和高斯核系数两个结构参数,据此建立滚动轴承故障分类模型;并利用美国凯斯西储大学轴承数据验证了方法的有效性。实验结果表明:与基于BP、SVM的故障分类模型相比,Hilbert包络谱奇异值和IPSO优化SVM的滚动轴承故障诊断分类模型具有更高的精度、更强的泛化能力。     

An insight concept to select appropriate IMFs for envelope analysis of bearing fault diagnosis

Traditional envelope analysis must examine all the resonant frequency bands during the process of bearing fault detection. To eliminate the above deficiency, this paper presents an insight concept based on the empirical mode decomposition to choose an appropriate resonant frequency band for characterizing feature frequencies of bearing faults by using the envelope analysis subsequently. By the band-pass filtering nature of the empirical mode decomposition, the resonant frequency bands are allocated in a specific intrinsic mode function. The inner or outer ring of bearings scratched intentionally is used to validate the feasibility of the proposed idea, and comparisons with the traditional envelope analysis are addressed. The experimental results show that the proposed insight concept can efficiently and correctly diagnose the bearing fault types.     

Bi-spectrum based-EMD applied to the non-stationary vibration signals for bearing faults diagnosis

Empirical mode decomposition (EMD) has been widely applied to analyze vibration signals behavior for bearing failures detection. Vibration signals are almost always non-stationary since bearings are inherently dynamic (e.g., speed and load condition change over time). By using EMD, the complicated non-stationary vibration signal is decomposed into a number of stationary intrinsic mode functions (IMFs) based on the local characteristic time scale of the signal. Bi-spectrum, a third-order statistic, helps to identify phase coupling effects, the bi-spectrum is theoretically zero for Gaussian noise and it is flat for non-Gaussian white noise, consequently the bi-spectrum analysis is insensitive to random noise, which are useful for detecting faults in induction machines. Utilizing the advantages of EMD and bi-spectrum, this article proposes a joint method for detecting such faults, called bi-spectrum based EMD (BSEMD). First, original vibration signals collected from accelerometers are decomposed by EMD and a set of IMFs is produced. Then, the IMF signals are analyzed via bi-spectrum to detect outer race bearing defects. The procedure is illustrated with the experimental bearing vibration data. The experimental results show that BSEMD techniques can effectively diagnosis bearing failures.     

关联维数和基于内禀模态函数的AR模型在滚动轴承故障诊断中的应用

提出一种基于内禀模态函数(Intrinsic Mode Function,IMF)、自回归(Auto-Regressive,AR)模型和关联维数的滚动轴承故障诊断方法.该方法首先采用经验模态分解(Empirical Mode Decomposition,EMD)将滚动轴承振动信号分解成若干个IMF,然后对包含主要故障信息的IMF分量建立AR模型,计算AR模型自回归参数的关联维数,并以关联维数作为特征向量输入神经网络分类器,最后通过网络的输出结果来识别轴承的工作状态和故障类型.对实验数据的分析结果表明,该方法能有效地应用于滚动轴承的故障诊断.     

基于峭度的VMD分解中k值的确定方法研究

在变分模态分解（Variationalmodedecomposition,VMD）中,因为分解层数k值属于自定义变量,所以在取值时分解结果会随着k值的变化而得到不同的结果,k值的取值直接影响着结果的准确性,k值取得过大或者过小都会对结果造成影响。基于以上问题,提出了一种利用峭度确定k值的方法。选取k值为2ⁿ的整数,计算当k值为2ⁿ时,每一个k值相关系数最大分量的峭度,并绘制峭度的变化曲线,若在该曲线内峭度没有峰值且单调递增,则继续计算当k为n+1时相关系数最大分量的峭度值,重复以上步骤,以峭度最大作为优化的标准,当峭度最大的时候,k为最佳值。用实际故障信号验证了方法的可行性,为VMD方法的研究提出了一种新的思路。     

基于EMD和ANFIS的自适应噪声消除研究

对于混入色噪声的混合信号,如果可以通过测量得到产生色噪声的白噪声,对白噪声进行非线性训练即可逼近色噪声,达到非线性滤波的目的.自适应模糊推理系统(adaptive neuro-fuzzy unference system,ANFIS)可以实现上述非线性逼近.文中在上述算法的基础上,提出一种EMD(empirical mode decomposition)-ANFIS的自适应色噪声消除方法,首先对混合信号进行EMD分解,得到各个内禀模态函数分量(intrinsic mode function, IMF),然后对分解得到的内禀模态分量进行ANFIS模糊消噪,最后对消噪后的各个分量信号进行叠加.由于所得内禀模态函数为近似平稳信号,且图形越来越趋于平缓,减小了ANFIS方法的逼近难度.在混合信号信噪比为2.840 7 dB时,经过EMD-ANFIS消噪后的估计误差比只经过ANFIS消噪后的估计误差减少11.74 dB,证明EMD-ANFIS方法的有效性.     

基于内蕴分量建模的滚动轴承异音信号分析新方法

针对滚动轴承振动信号的非平稳时变特性,采用局域波和K-L信息量的分析方法首先把轴承的波形图通过局域波分解为多个内蕴分量,然后对参考波形图的内蕴分量以及待测波形图的内蕴分量进行K-L信息量的自回归(Autoregressive, AR) 建模,求出它们之间的残差方差,通过人工试验的方法来不断的修正AR模型,直到AR模型满足检测的准确率。通过以正常状态为参考状态,对滚动轴承实例在三种不同状态下进行了分析比较试验,结果证明该方法在滚动轴承异音探测以及分析方面可以达到用户预先设定的高准确度,具有很高的工程实用性。     

剥落形状对齿轮啮合刚度及动力学特性的影响

传统的剥落齿故障研究主要探讨剥落齿在不同剥落深度和不同剥落宽度的情况下齿轮刚度变化,具有较大的局限。针对此问题,对剥落形状、剥落形状分布对齿轮刚度和动力学的影响进行了研究。首先,取剥落图形在分度圆上,在此条件下取不同剥落图形。然后,在传统势能法基础上,考虑齿轮啮合位置变化时剥落宽度变化,以此为基础求出不同剥落图形的刚度变化曲线;再结合动力学变化,对于不同故障振动加速度信号进行分析,说明剥落形状因素影响齿轮相关指标,使得剥落齿轮与正常齿轮难以判别,指标相互混叠。提出使用概率密度函数与峭度值相结合的方法,利用概率密度函数曲线判断齿轮故障状态。发现此种方法可以有效抵消形状因素对于指标的干扰,具备较高可行性及判别能力。     

基于VMD和对称差分能量算子解调的滚动轴承故障诊断方法

针对滚动轴承早期故障振动信号非平稳、强噪声,故障频率难提取的问题,提出了基于变分模态分解（VariationalModeDecomposition,VMD）和对称差分能量算子解调的滚动轴承故障诊断方法。首先,利用VMD方法将滚动轴承待分析信号分解成若干个模态分量;其次,根据峭度最大准则来选取被对称差分能量算子解调的模态分量,解调后获取待分析信号的幅值、频率信息并计算包络谱。实验结果表明:与传统能量算子相比,所提方法能突显故障特征频率并有效抑制虚假干扰频率,更有利于滚动轴承故障诊断。     

基于自适应EEMD与Fast Kurtogram的齿轮箱故障诊断

针对集合经验模式分解算法中添加白噪声幅值大小和总体平均次数过分依赖于人的主观经验或多次尝试,具有较大主观性和盲目性的不足,提出一种自适应EEMD结合快速峭度图(Fast Kurtogram)的故障诊断方法。首先将采集到的振动信号进行EMD预处理以自适应的获取EEMD算法的关键输入参数,然后结合峭度与互相关系数“双阀值准则”快速选取分量进行信号重构以突出故障特征,并通过快速峭度图选取最佳滤波参数,最后对滤波后的信号做包络谱分析,实现故障特征频率的提取与故障诊断。通过模拟信号分析及减速器齿轮箱的故障诊断工程应用,并与EMD方法及传统EEMD方法进行对比分析,验证了提出方法的有效性。结果表明,所提出的方法能够从含有强烈背景噪声的信号中成功提取出减速器齿轮箱的早期微弱故障特征,提高了故障诊断的及时性与准确性。     

Application of computed order tracking, Vold-Kalman filtering and EMD in rotating machine vibration

This paper presents a study on rotating machine vibration signals by using computed order tracking, Vold-Kalman filtering and intrinsic mode functions from the empirical mode decomposition method. Through the sequential use of intrinsic mode function and order tracking methods, both speed synchronous and non-synchronous vibrations that modulate orders in rotating machine vibrations are distinguished, which is difficult when using each of the techniques in isolation alone. Simulation and experimental studies demonstrate the ability of extracting vibrations that modulate order signals through combining the techniques.     

Scanning assembly

A new solution to the problem of assembling chamferless parts, the scanning assembly method (SAM), has been developed. A SAM-based device acts as a multiaxis microrange float connected to the end of a robot arm. It consists of a remote centre compliance device and a vibrator, and imposes high-frequency vibrational motion on one of the parts to be assembled. The induced vibrations are harmonic functions of time, and take place along and around two perpendicular axes normal to the axis of insertion. These vibrations and the compliance of the SAM-based device allow any linear and angular misalignments to be overcome continuously. The device has been used experimentally to assemble chamferless pegs into chamferless holes.     

An efficient algorithm for processing data from a multi-point sensor of vibrations

A new technique of vibration sensing, based on a polarimetric fibre-optic strain sensor, is presented; it is designed for localisation of multiple sources of disturbances in a broad frequency spectrum without using fibre gratings. A mathematical model of the sensor is used for development of a variational method for estimation of amplitudes of component vibrations on the basis of noisy samples of the signal at the output of the sensor. This method is implemented in a new algorithm of estimation, being 100–1000 more efficient (in terms of computing time) than an algorithm published previously.     

The Huang Hilbert Transform for evaluating the instantaneous frequency evolution of transient signals in non-linear systems

Traditional approaches for the analysis of transient signals are generally based on the apriori knowledge of the system under test for choosing a preliminary set of waveforms; consequently, they use a mathematical algorithm to decompose the signal itself into a suitable combination of the chosen waveforms. Conversely, this work is aimed to investigate the possibility of extracting the features of transient signals through the evaluation of their instantaneous frequency evolution. For this aim, the Huang Hilbert Transform (HHT) has been exploited (i) to decompose the input signal into a set of Intrinsic Mode Functions (IMFs), (ii) to extract the IMFs analytical signals, (iii) to evaluate their amplitude and phase evolutions, (iv) to compute the instantaneous frequency of the input signal and (v) to extract the signal information searched for. In order to evaluate its performance, the proposed approach has been firstly applied to a synthesized signal with known instantaneous amplitude and frequency evolution. Successively, in order to assess the reliability of HHT results with signals acquired on experimental circuits, the current flowing in an actual RLC circuit during its free natural oscillation has been analyzed. With the aim of analyzing the performance gained also in the presence of evident non-linearities, a saturable inductor has been introduced in the test circuit. Also in this case, by comparing the achieved results with those shown by different traditional approaches, great advantages have been experienced in terms of accuracy. Furthermore, beyond the accurate frequency representation, the experimental results evidenced the intrinsic ability of the proposed approach to extract meaningful information related to the knowledge of the underlying process. Finally, it is worth noting that the results reported in this paper requested no apriori knowledge about the signal/process under test.