基于改进EMD与PNN的汽轮机通流部分故障诊断

根据热力参数非线性、非稳态的特点,提出了一种基于改进的经验模态分解(empirical mode decomposition,简称EMD)算法与概率神经网络(probabilistic neural network,简称PNN)的汽轮机通流部分故障诊断新方法。该方法针对EMD存在的端点效应问题,采取基于波形相似度的镜像延拓法进行改进,以得到更准确、更真实的本征模函数(intrinsic mode function,简称IMF)分量,从而有效提取了故障特征信息,并通过PNN训练判别汽轮机通流部分故障类型。以某电厂600MW火电机组实时运行数据为基础进行仿真实验,结果表明,基于改进EMD与PNN的汽轮机通流部分诊断方法能够快速准确地判别汽轮机通流部分的故障类型,其准确率明显高于基于EMD与PNN的故障诊断方法。     

基于CEEMDAN模糊熵和SVM的滚动轴承故障诊断

针对滚动轴承振动信号非平稳性及特征信息复杂问题,提出一种基于自适应白噪声完备经验模态分解模糊熵和支持向量机相结合的故障诊断方法。首先,采用CEEMDAN,将滚动轴承振动信号分解成若干个固有模态函数分量,并根据相关系数—峭度准则选取可有效表征信号自身特性的模态分量;然后,计算各敏感IMF分量模糊熵,并构建高维特征向量;最后,将高维特征向量输入SVM中,以实现对故障类型和工作状态的有效诊断。试验结果表明,该方法可有效对滚动轴承故障进行诊断。     

基于EMD和SVM的刀具故障诊断方法

为了解决刀具在切削过程中出现的故障,提出了基于经验模态分解(EMD)和支持向量机(SVM)的刀具故障诊断方法.该方法首先将经过标准化的声发射信号进行经验模态分解,将其分解为有限个固有模态函数(IMF)和残余量之和,然后对每个IMF分量通过一定的削减算法增强故障类型特征,计算每个IMF和残余项的能量以及整个信号的削减比作...     

基于CEEMDAN奇异值熵和SVM的转子故障诊断

提出一种基于自适应白噪声完整经验模态分解(complete ensemble empirical mode decomposition with adaptive noise,CEEMDAN)奇异值熵和支持向量机(support vector machine,SVM)的转子故障诊断方法。利用CEEMDAN方法首先对非平稳的转子振动信号分解得到若干个表征信号自身特性的固有模态函数(intrinsic mode function,IMF),并通过虚假IMF分量判别法,剔除对于故障特征不敏感的IMF,以保证故障信息提取的准确性和有效性,在此基础上产生初始特征向量矩阵。并对此矩阵进行奇异值分解得到矩阵奇异值,使其作为故障特征向量,通过归一化处理得到奇异值熵,并以此作为SVM的输入,对转子的工作状态进行识别。研究结果表明:该方法可有效应用于转子故障诊断,实现对转子工作状态和故障类型的有效诊断。     

基于多元经验模态分解互近似熵及GG聚类的轴承故障诊断

提出了一种基于多元经验模态分解（Multi-EMD）、互近似熵和GG聚类的滚动故障轴承诊断方法。首先,将振动信号进行多元经验模态分解,得到若干个内禀模态函数（IMF）分量和一个趋势项。然后,将IMF分量分别与原始信号进行相关性分析,筛选出前7个含主要特征信息的IMF分量,并将筛选的IMF分量的互近似熵作为特征向量。最后,将特征向量输入到GG模糊分类器中进行聚类识别。通过聚类三维图,对两种算法机械运行的4种状态进行了对比,验证了多元经验模态分解方法不仅可解决采样的不均衡问题,而且可解决EMD算法聚类的混叠问题。     

基于改进EMD样本熵和SVM的风机滚动轴承故障诊断

风机齿轮箱振动信号成分复杂,而经验模态分解(EMD)在故障诊断中存在模态混叠和端点效应问题.针对此问题,研究了一种EEMD样本熵和高斯径向基核函数的SVM分类器的滚动轴承故障诊断方法.以风机齿轮箱滚动轴承为研究对象,提取了内圈故障、外圈故障、滚动体故障和正常轴承4种状态振动信号,利用EEMD和小波分别对振动信号分解降噪并筛选主要IMF分量;计算前4阶IMF分量的样本熵作为特征向量;最后将特征向量输入高斯径向基核函数的SVM模型进行故障识别.结果表明:EEMD算法对端点效应和模态混叠都有一定抑制作用,EEMD样本熵和SVM相结合可有效识别滚动轴承故障类型,故障识别率为97.5％,为工程应用中风机齿轮箱滚动轴承故障诊断提供参考.     

自适应的EEMD及其在滚动轴承故障诊断中的应用

为解决总体集成经验模态分解(EEMD)算法中存在的白噪声参数需要人为选择的问题,同时考虑到现实中难以获得大量典型故障样本的实际情况,提出了一种基于自适应总体集成经验模态分解(AEEMD)与支持向量机(SVM)的滚动轴承故障诊断法。首先在信号处理上使用AEEMD将原始振动信号分解成具有不同特征时间尺度的本征模态分量(IMF),对于不同的轴承故障来说,在不同频带内的信号能量会发生改变,因此可通过计算各个IMF的能量来实现故障特征提取;然后把IMF的能量特征值作为输入来构建支持向量机分类器模型;最后利用建立的模型对轴承的状态类型做出判别。在轴承故障实例中将AEEMD算法与EEMD算法进行对比,证明了AEEMD的分解效果更好;选用BP神经网络与SVM的诊断效果进行对比分析,表明本文中提出的方法能够更加快速准确地诊断出轴承的故障。     

基于MCKD和改进HHT多尺度模糊熵的齿轮故障诊断方法

针对齿轮故障信号常伴有大量噪声,故障特征难以提取的问题,提出一种基于最大相关峭度解卷积(MCKD)和改进希尔伯特-黄变换(HHT)多尺度模糊熵的故障诊断方法。首先采用MCKD算法对采集到的齿轮振动信号进行降噪处理,以提高信号的信噪比;然后利用自适应白噪声完备经验模态分解(CEEMDAN)对降噪后信号进行分解,获得一系列不同尺度的固有模态函数(IMF),并通过相关系数-能量的虚假IMF评价方法选取对故障敏感的模态分量;最后计算敏感IMF分量的模糊熵,将获得的原信号多尺度的模糊熵作为状态特征参数输入最小二乘支持向量机(LS-SVM)中,对齿轮的故障类型进行诊断。实测信号的诊断结果表明,该方法可实现齿轮故障的有效诊断。     

IMF能量和RBF神经网络相结合在滚动轴承故障诊断中的应用研究

针对滚动轴承的故障特点,提出了一种将IMF能量与RBF神经网络相结合的方法用于故障诊断.该方法首先利用经验模态分解(EMD)方法,把振动信号分解为若干个IMF分量,再用重要的IMF分量求得IMF能量特征向量,最后将特征向量输入RBF神经网络进行故障模式分类.通过对滚动轴承的正常状态、内圈故障、滚动体故障和外圈故障信号的分析结果表明,该方法能够准确、有效地识别这些故障.     

CEEMDAN辅助快速谱峭度图算法的轴承故障诊断

针对滚动轴承的振动信号受到强噪声污染故障特征提取困难的问题,提出一种结合完整集成经验模态分解(CEEMDAN)辅助快速谱峭度图算法(Fast Kurtogram)的轴承故障诊断方法,首先将信号通过CEEMDAN分解为多个模态分量(IMF)及余量,计算出各个IMF分量的峭度值,并利用峭度值对分量信号进行重构;然后利用快速谱峭度图算法对重构信号进行处理,可以得到良好的故障带宽与中心频率。实测数据表明,该方法能够有效检测出轴承内圈故障。     

Application of EEMD and improved frequency band entropy in bearing fault feature extraction

Ensemble empirical mode decomposition (EEMD) is widely used in condition monitoring of modern machine for its unique advantages. However, when the signal-to-noise ratio is low, the de-noising function of it is often not ideal. Thus, a new fault feature extraction method for rolling bearing combining EEMD and improved frequency band entropy (IFBE) is proposed, i.e., EEMD–IFBE. According to the problem of multiple intrinsic mode functions (IMFs) generated by EEMD, how to select the sensitive IMF(s) that can better reflect fault characteristics, a novel method based on FBE for sensitive IMF is proposed. In addition, since the bandwidth parameter is set empirically when the band-pass filter is designed based on the original FBE, a novel bandwidth parameter optimization method based on the principle of maximum envelope kurtosis is proposed. First, the original vibration signal is subjected to EEMD to obtain a series of IMFs; Then, the FBE values are obtained for the original signal and each IMF component, and the bandwidth of the band-pass filter (empirically) is designed as the characteristic frequency band at the minimum entropy value, and the affiliation between the characteristic frequency band of each IMF and the characteristic frequency band of the original signal is compared, and then selecting the sensitive IMF(s) that reflects the characteristics of the fault; Third, due to the influence of background noise, it is difficult to accurately obtain the fault frequency from the selected IMF(s). Therefore, the band-pass filter designed based on FBE is used, and the bandwidth parameter is optimized based on the principle of envelope kurtosis maximum, and then the selected sensitive IMF is band-pass filtered. Finally, the envelope power spectrum analysis is performed on the filtered signal to extract the fault characteristic frequency, and then the fault diagnosis of the bearing is realized. The method is successfully applied to simulated data and actual data of rolling bearing, which can accurately diagnose fault characteristics of bearing and prove the effectiveness and advantages of the method.     

HHT与Elman神经网络在离心泵故障振动信号处理中的应用

提出了一种Hilbert-Huang变换(HHT)和Elman神经网络相结合的离心泵振动信号故障诊断新方法.首先,将离心泵振动信号时间序列数据经验模态分解(Empirical Mode Decomposition,简称EMD),然后经过Hilbert-Huang变换获得各模态(Intrinsic Mode Functions,简称IMF)的能量,并以"能量比"为元素,构造离心泵振动信号的特征向量,根据Elman神经网络模型能够逼近任意非线性函数的特点和具有反映系统动态特性的能力,利用Elman神经网络模型实现离心泵故障的诊断.实验研究结果表明该方法对离心泵振动信号故障具有很高的诊断率.     

EEMD和TFPF联合降噪法在齿轮故障诊断中的应用

为了消除噪声对齿轮传动系统故障特征提取的影响,提出了一种基于集成经验模态分解(ensemble empirical mode decomposition,简称EEMD)和时频峰值滤波(time-frequency peak filtering,简称TFPF)相结合的降噪方法。针对TFPF算法在窗长的选择方面受到限制的问题,采用了EEMD方法对其进行改进,使得信号在噪声压制和有效信号保真两方面得到权衡;含噪声的信号经过EEMD分解后,得到一系列频率成分从高到低的本征模态函数(intrinsic mode functions,简称IMFs),计算出各IMFs间的相关系数,判断需要滤波的IMFs。对不同的IMFs选择不同的窗长进行TFPF滤波,把过滤后的IMFs和剩余的IMFs重构得到最终的降噪信号。用模拟仿真信号和齿轮齿根故障信号对该方法进行验证,可见EEMD+TFPF能有效地去除噪声,成功提取齿根裂纹故障特征。     

基于本征模态函数包络谱的滚动轴承故障诊断

针对滚动轴承故障振动信号非平稳的特征,以及传统傅里叶变换不能反映信号细节的缺陷,引入了一种基于本征模态函数包络谱的方法。首先,采用经验模态分解（empirical mode decomposition,EMD）将滚动轴承故障振动信号分解成若干个本征模态函数（intrinsic mode function,IMF）之和;然后,求出包含主要信息成分的IMF分量的Hilbert包络谱;最后,对照滚动轴承故障特征频率,进而判定故障类型。通过对滚动轴承内圈、外圈故障振动信号的分析处理,表明该方法能有效地提取滚动轴承的故障特征。     

A fault diagnosis approach for roller bearing based on IMF envelope spectrum and SVM

Targeting the modulation characteristics of roller bearing fault vibration signals, a method of fault feature extraction based on intrinsic mode function (IMF) envelope spectrum is proposed to overcome the limitations of conventional envelope analysis method. By utilizing the proposed feature extraction method, the disadvantages of conventional envelope analysis method such as the chosen of central frequency of filter with experience in advance, looking for spectral line of fault characteristic frequencies in envelope spectrum and so on could be overcome. Firstly, the original modulation signals are decomposed into a number of IMFs by empirical mode decomposition (EMD) method. Secondly, the ratios of amplitudes at the different fault characteristic frequencies in the envelope spectra of some IMFs that include dominant fault information are defined as the characteristic amplitude ratios. Finally, the characteristic amplitude ratios serve as the fault characteristic vectors to be input to the support vector machine (SVM) classifiers and the work condition and fault patterns of the roller bearings are identified. Since the recognition results are available directly from the output of the SVM classifiers, the proposed diagnosis method provides the possibility to fulfill the automatic recognition to machinery faults.     

热力参数的KPCA-RBF网络建模及传感器故障诊断方法

针对复杂恶劣环境下机组热力参数的数据监测及传感器故障诊断问题,建立了融合机理分析、核主元分析(kernel principle component analysis,简称KPCA)与径向基神经网络(radial basis function,简称RBF)的发电机组热力参数预测及传感器故障检测模型。首先,根据机理分析得到完备的辅助变量集,并利用核主元分析提取辅助变量的特征信息以有效处理发电机组中高维、强耦合的非线性数据;其次,将主元变量集输入径向基神经网络进行学习,实现热力参数的重构;最后,基于预测模型与窗口移动法实现传感器的故障诊断,并对故障数据进行及时修复和准确替换。以燃气轮机排气温度为例进行验证的结果表明,该预测模型具有更高的精度和泛化能力,能在传感器故障发生初期及时发现并识别故障类型,检测效果优良。     

基于残差分析的挠度传感器故障时间定位算法

及时诊断出桥梁健康监测系统中传感器故障及确定传感器发生故障的精确时刻,可以有效地降低系统虚假报警率、保障系统正常运行。通过分析桥梁传感器故障演变过程,建立了数据标准化残差模型,在该模型基础上提出了一种基于数据标准化残差分析的故障时间定位算法,该算法以传感器采集数据的残差偏离量化值对故障进行判定,从而对传感器故障时间进行定位。以系统中常见的挠度传感器作为研究对象,针对工程中常见的常值故障、固定偏差、精度下降、漂移故障4种故障类型进行了仿真实验。实验结果表明,该方法对挠度传感器4种故障的时间定位精度有较大幅度的提高。     

基于优化自适应阈值的非线性机电系统传感器 故障检测和主动容错控制

针对具有参数不确定性和传感器故障的非线性机电系统,提出一种基于优化自适应阈值和故障重构策略的主动容错控制方法。首先,利用线性分式变换理论对存在参数不确定性的非线性机电系统进行建模,并提出基于粒子群优化算法的优化自适应阈值以提高参数不确定条件下的故障检测性能。其次,通过解析冗余关系推导出系统的动力学方程,并提出一种基于递归终端滑模的跟踪控制策略,以实现系统健康状态下的负载位置跟踪。当系统发生故障时,构建自适应滑模观测器进行传感器故障重构,根据重构结果设计自适应主动容错控制律,并利用故障检测结果进行控制律的实时切换。实验结果表明,所提出的故障检测和主动容错控制方法能在0.06 s内准确的实现传感器故障检测和容错控制,验证了该方法的可行性。     

Fault diagnosis method for rolling bearing based on VMD and improved SVM optimized by METLBO

Vibration signal processing and classification are critical for bearing fault diagnosis. In this study, a hybrid framework based on multi-envelopment teaching-learning-based optimization (METLBO) was proposed by combining parameters optimized variational mode decomposition (VMD) and improved support vector machines (ISVM). First, the average value of minimum enveloping entropy was considered the objective function of the optimizer, and the optimal parameters of VMD were obtained through METLBO optimization. Next, these optimal parameters were adopted to decompose the fault signal into intrinsic modal functions (IMFs). For ensuring fault feature robustness, the eigenvectors were formed by the energy and envelope entropy of IMFs. Finally, the ISVM model was established for training and testing by adding an input layer to the SVM to perform soft thresholding on input data. METLBO was adopted to determine the optimal soft threshold values of features and hyper-parameters of ISVM. The experimental comparison analysis revealed the effectiveness of the proposed method for bearing fault diagnosis.