Adaptive modelling of transient vibration signals

Conventional vibration signal processing techniques are most suitable for stationary processes. However, most mechanical faults in machinery reveal themselves through transient events in vibration signals. Time-series modelling, including autoregressive moving average (ARMA) modelling and autoregressive (AR) modelling, is an efficient approach for transient signal analysis. Based on the adaptive prediction technique, this paper applies the principle of the adaptive line enhancer (ALE) to the modelling of transient vibration signals. The time-series models, adaptive algorithms and the rational time–frequency transfer function are investigated in the paper. Simulation and experimental studies with different time–frequency–amplitude distributions and transient vibration responses are described. The results show that the adaptive modelling method can trace the time–frequency signal and extract dynamic features such as time–frequency distributions and time–amplitude distributions from sample signals. Given the simple programming and potentially easy implementation in on-line applications, this method should have application in machine monitoring and fault diagnosis.     

基于自适应时频分解的瞬态冲击信号提纯

在分析非平稳振动信号中瞬态冲击信号和平稳振动信号各自特点的基础上,发展了新的提纯机械故障冲击成分的信号处理方法。首先选择Chirplet基函数将被分析信号自适应展开,得到噪声抑制,高分辨率且无交叉干扰项的自适应时频谱,然后分别使用了两种方法从自适应时频谱中恢复出瞬态冲击分量。一种是采用振动冲击信号模型的时频分布去逼近时频滤波后的冲击分量时频分布,继而借助信号模型重构出冲击分量;另一种是根据冲击信号的Chirplet基函数的参数表达特征,直接选择所需基函数对冲击分量进行重构。最后,采用这两种方法分别对齿轮和轴承的故障信号进行分析,分析结果验证了这两种方法对瞬态冲击信号提取都非常有效,在机械故障诊断中有很大的参考价值。     

齿轮振动信号分解及其在故障诊断中的应用

对齿轮振动信号的测试及分解进行了研究。根据信号基频,把齿轮振动信号分解为啮合振动与旋转振动,这些振动信号可用于对齿轮状态进行定量研究。基于不同形式的齿轮振动信号,介绍了几种方法来提取信号中的故障信息。利用时域平均技术及齿轮振动信号分解理论对某齿轮箱早期故障信号进行了检测。研究表明,齿轮运动信号分解能够有效检测齿轮的各类故障,高阶加速度信号对齿轮某些类型的早期故障更加敏感。     

Estimation of modal parameters using bilinear joint time–frequency distributions

In this paper, a new method is proposed for modal parameter estimation using time–frequency representations. Smoothed Pseudo Wigner–Ville distribution which is a member of the Cohen's class distributions is used to decouple vibration modes completely in order to study each mode separately. This distribution reduces cross-terms which are troublesome in Wigner–Ville distribution and retains the resolution as well. The method was applied to highly damped systems, and results were superior to those obtained via other conventional methods.     

Detection of signal transients based on wavelet and statistics for machine fault diagnosis

This paper presents a transient detection method that combines continuous wavelet transform (CWT) and Kolmogorov–Smirnov (K–S) test for machine fault diagnosis. According to this method, the CWT represents the signal in the time-scale plane, and the proposed “step-by-step detection” based on K–S test identifies the transient coefficients. Simulation study shows that the transient feature can be effectively identified in the time-scale plane with the K–S test. Moreover, the transients can be further transformed back into the time domain through the inverse CWT. The proposed method is then utilized in the gearbox vibration transient detection for fault diagnosis, and the results show that the transient features both expressed in the time-scale plane and re-constructed in the time domain characterize the gearbox condition and fault severity development more clearly than the original time domain signal. The proposed method is also applied to the vibration signals of cone bearings with the localized fault in the inner race, outer race and the rolling elements, respectively. The detected transients indicate not only the existence of the bearing faults, but also the information about the fault severity to a certain degree.     

机械振动微弱慢频变信号的混沌振子检测

机械振动微弱信号的检测与识别有利于早期故障的检测与诊断。用Duffing混沌振子检测微弱振动信号具有明显的优势。提出了用混沌振子检测慢变频微弱振动信号的方法。在给出Duffing混沌振子对微弱信号检测的基本原理后，根据慢频变信号的特征，对信号进行了周期离散，提出了对暂态信号进行时域延拓的方法，分析了可行性。提出了幅频联调方法，设计了检测原理，并给出了实现步骤。结合所提出的两种方法，对带噪声的频变微弱振动信号进行了检测分析。仿真结果和实际采集信号分析结果支持了所提出方法的适用性。     

旋转机械振动信号的信息熵特征

从信息融合的思想出发 ,针对单个和多个振动传感器 ,在时域、频域以及时 -频域系统、深入地研究了定量评价旋转机械振动状态的方法 ,提出了反映不同域中振动能量分布不确定性的奇异谱熵、功率谱熵、涡动状态特征熵、小波空间特征熵等信息熵特征。通过对实际信号的分析表明 ,这些信息熵形成了有效综合评价转子振动状态的特征指标。     

Bearing fault detection utilizing group delay and the Hilbert-Huang transform

Vibration signals measured from a mechanical system are useful to detect system faults. Signal processing has been used to extract fault information in bearing systems. However, a wide vibration signal frequency band often affects the ability to obtain the effective fault features. In addition, a few oscillation components are not useful at the entire frequency band in a vibration signal. By contrast, useful fatigue information can be embedded in the noise oscillation components. Thus, a method to estimate which frequency band contains fault information utilizing group delay was proposed in this paper. Group delay as a measure of phase distortion can indicate the phase structure relationship in the frequency domain between original (with noise) and denoising signals. We used the empirical mode decomposition of a Hilbert-Huang transform to sift the useful intrinsic mode functions based on the results of group delay after determining the valuable frequency band. Finally, envelope analysis and the energy distribution after the Hilbert transform were used to complete the fault diagnosis. The practical bearing fault data, which were divided into inner and outer race faults, were used to verify the efficiency and quality of the proposed method.     

Multiscale singular value manifold for rotating machinery fault diagnosis

Time-frequency distribution of vibration signal can be considered as an image that contains more information than signal in time domain. Manifold learning is a novel theory for image recognition that can be also applied to rotating machinery fault pattern recognition based on time-frequency distributions. However, the vibration signal of rotating machinery in fault condition contains cyclical transient impulses with different phrases which are detrimental to image recognition for time-frequency distribution. To eliminate the effects of phase differences and extract the inherent features of time-frequency distributions, a multiscale singular value manifold method is proposed. The obtained low-dimensional multiscale singular value manifold features can reveal the differences of different fault patterns and they are applicable to classification and diagnosis. Experimental verification proves that the performance of the proposed method is superior in rotating machinery fault diagnosis.     

基于双时域微弱故障特征增强的轴承早期故障智能识别*

针对轴承早期微弱故障难以准确识别的问题,提出一种基于双时域微弱故障特征增强的轴承早期故障智能识别方法。利用广义S变换和Fourier逆变换推导出一种双时域变换,将轴承振动信号变换为双时域二维时间序列。根据双时域变换的能量分布特点,提取二维时间序列的主对角元素以构建故障特征增强的时域振动信号。仿真信号和轴承故障信号分析验证了双时域微弱故障特征增强的可行性和有效性。采用脉冲耦合神经网络和支持向量机对增强后的轴承信号进行时频特征参数提取和智能识别,平均识别精度达到了95.4%。试验结果表明所提方法能有效提高轴承早期故障的智能识别精度。     

Application of time–frequency entropy method based on Hilbert–Huang transform to gear fault diagnosis

When faults occur in the gear, energy distribution of gear vibration signals measured in time–frequency plane would be different from the distribution under the normal state. Therefore, it is possible to detect a fault by comparing the energy distribution of gear vibration signals with and without fault conditions. Hilbert–Huang transform can offer a complete and accurate energy–frequency–time distribution. On the other hand, Shannon entropy could give a useful criterion for analyzing and comparing probability distribution and offer a measure of the information of any distribution. Targeting the feature of energy distribution of gear vibration signal, the merit of entropy and Hilbert–Huang transform, the concept of time–frequency entropy based on Hilbert–Huang transform is defined and furthermore gear fault diagnosis method based on time–frequency entropy is proposed. The analysis results from simulated signals and experimental signals with normal and defective gears show that the diagnosis approach proposed could identify gear status-with or without fault accurately and effectively. However, further study is needed to the classify gear fault pattern such as crack fault or broken teeth.     

一种频域特征提取自编码器及其在故障诊断中的应用研究

提出一种可以直接从振动信号中提取频域特征的非对称自编码器方法。与传统自编码器以重构振动信号作为目标输出不同,频域自编码器使用振动信号的频谱作为目标输出,这种非对称的自编码器可以学习振动信号与其频谱之间的映射关系,使得编码器可以输出频域特征。为了说明提出的频域自编码器的特征提取效果,在轴承数据集上进行特征提取和故障诊断实验,在没有引入标签信息的情况下,频域自编码器提取到的特征表现出较好的聚类效果,能够区分轴承的不同故障类型;进一步进行了泛化实验,训练分类器时使用1%的有标签样本,可以达到90%以上的故障分类准确率。实验结果表明,频域自编码器与传统自编码器相比,可以更好地提取振动信号的故障特征信息,具有一定的实用价值。     

小波包分析在轴承早期故障诊断中的应用

为了识别轴承早期损伤引起的故障信号,利用小波包对轴承的振动信号进行处理。小波包分析的实质是对小波分解的结果作进一步细分,因而具有比小波分解高得多的频域分辨能力。文中用小波包分析了两个存在早期轻微损伤的轴承的振动信号,并比较了自然序、Gray序以及移频算法的处理结果。这些分析结果表明,小波包分析能够有效地将隐藏在正常振动信号之中的早期弱故障信号提取出来,从而发现轴承的早期损伤。     

某型燃气涡轮起动机振动测试系统的研发

振动信号能够从时域或频域实时地反映旋转机械的故障信息,为满足某型燃气涡轮起动机故障诊断的需要,研发了涡轮起动机振动测试系统。该系统能够控制涡轮起动机的工作过程,监控工作状态,记录工作数据和分析振动信息。系统以起动机时域振动信息为基础,运用频谱分析和倒谱分析方法处理振动信息,以确定故障部位。系统研究了燃气涡轮起动机与加载系统匹配的方法,以及加载系统附加振动信号的识别与分离。试验结果表明,该系统能够实现涡轮起动机工作过程的控制,工作数据和振动信息的采集和分析,并且可以识别与分离加载系统的附加振动信号,为燃气涡轮起动机故障诊断提供可靠的实验研究平台。     

几种Cohen类时频分布的比较及应用

比较了谱图、平滑伪Wigner-Ville分布、Choi-Williams分布和Cone核分布四种Cohen类时频分布方法及其性能,从理论上讨论了交叉项衰减和时频聚集度的性能,并分析了相应的其他特性。针对齿轮箱状态监测问题,通过仿真和试验信号,展现了上述方法的应用效果。结果表明Choi-Williams分布在非平稳信号所引发的短时变化中具有高度的敏感性,更为适合表征非平稳信号的时变信息,易于诊断出齿轮的局部损伤故障,尤其适用于故障的早期诊断。     

小波包分析方法在齿轮早期故障特征提取中的应用

基于小波包对信号的高分辨率分解和重构能力,把信号分解到不同频段,然后选择有效频段进行故障信号重构,分离出故障信息,试验表明,该方法能从很强的总体振动信号中提取清晰的损伤特征,实现早期诊断。     

Fault Feature Extraction of Diesel Engine Based on Bispectrum Image Fractal Dimension

Fault feature extraction has a positive effect on accurate diagnosis of diesel engine. Currently, studies of fault feature extraction have focused on the time domain or the frequency domain of signals. However, early fault signals are mostly weak energy signals, and time domain or frequency domain features will be overwhelmed by strong back?ground noise. In order consistent features to be extracted that accurately represent the state of the engine, bispectrum estimation is used to analyze the nonlinearity, non?Gaussianity and quadratic phase coupling(QPC) information of the engine vibration signals under different conditions. Digital image processing and fractal theory is used to extract the fractal features of the bispectrum pictures. The outcomes demonstrate that the diesel engine vibration signal bispectrum under different working conditions shows an obvious differences and the most complicated bispectrum is in the normal state. The fractal dimension of various invalid signs is novel and diverse fractal parameters were utilized to separate and characterize them. The value of the fractal dimension is consistent with the non?Gaussian intensity of the signal, so it can be used as an eigenvalue of fault diagnosis, and also be used as a non?Gaussian signal strength indicator. Consequently, a symptomatic approach in view of the hypothetical outcome is inferred and checked by the examination of vibration signals from the diesel motor. The proposed research provides the basis for on?line monitoring and diagnosis of valve train faults.     

Time–frequency analysis of friction-induced vibration under reciprocating sliding conditions

A time–frequency analysis can give an overall view of the behaviour of friction-induced vibration. In this paper, short-time Fourier transform (STFT), Wigner–Ville distribution (WVD), Choi–Williams distribution (CWD) and Zhao–Atlas–Marks distribution (ZAMD) are applied to analyze time–frequency characteristics of friction-induced vibration. The result shows that there is always a frequency change in the time–frequency presentation of vibration in the location where the vibration is bounded. The frequency changes in time–frequency presentations are associated with nonlinearity of vibration systems. The nonlinearity may be counted as the evidence to support the consideration that friction-induced vibrations are bounded by limit cycles due to the system nonlinearity. Based on the time–frequency presentations of vibrations, it may be concluded that the friction vibration system is generally a linear system in the phase of vibration initiation but is a nonlinear system in the phases of vibration being bounded and disappearance.     

Diagnosis of subharmonic faults of large rotating machinery based on EMD

The vibration signals always carry the abundant dynamic information of a machine and are very useful for the feature extraction and fault diagnosis. In practice, most subharmonic signals have a close relationship to time variables and can manifest large amplitude fluctuation, transient vibration, or modulation signals in time domain. In view of this, this paper describes an effective method to search the features of subharmonic faults of large rotating machinery based on empirical mode decomposition (EMD). Case study on some actual vibration signals of machine parts shows that EMD is an adaptive and unsupervised method in feature extraction and it provides an attractive alternative to the traditional diagnostic methods.     

高速列车转向架区气动噪声分离研究

提出了一种基于快速路径优化的自适应短时傅里叶变换时频分析方法,并将该方法用于行星齿轮箱的故障诊断。该时频分析方法通过使用快速路径优化获得瞬时频率变化规律,在短时傅里叶变换过程中自适应的改变时窗长度,从而获得更恰当的时频分辨率。针对行星齿轮箱运行状态不稳定的特点,通过使用笔者提出的时频分析方法可以有效地提取出行星齿轮箱的转速信息,利用参考转速对故障信号角度域重采样和阶次分析,从而实现变转速情况下的行星齿轮箱故障诊断。仿真分析表明,与传统短时傅里叶变换相比基于快速路径优化的自适应短时傅里叶变换得到的时频分布能量更加集中;试验分析证明了基于快速路径优化的自适应短时傅里叶变换方法在行星齿轮箱故障诊断中的有效性。