基于改进经验小波变换的行星齿轮箱故障诊断

行星齿轮箱振动信号具有复杂多分量和调幅-调频的特点。幅值解调和频率解调方法能够避免传统Fourier频谱中的复杂边带分析,有效识别故障特征频率。经验小波变换通过对信号Fourier频谱的分割构造一组正交滤波器组,能提取具有紧支撑Fourier频谱的单分量成分,再对单分量成分运用Hilbert变换即可实现信号的解调分析。经验小波变换能够有效分离出调幅-调频成分,不存在模态混叠现象,具有完备的理论基础,自适应性好、算法简单、计算速度快。将改进的经验小波变换应用于行星齿轮箱振动信号的解调分析;提出了一种单分量个数的估算方法,解决了经验小波变换中的Fourier频谱划分问题;给出了对故障敏感的信号分量的选取方法,提高了分析的针对性。将改进方法应用于行星齿轮箱振动仿真信号和实验信号分析,验证了该方法的有效性。     

层次朴素贝叶斯分类器构造算法及应用研究

贝叶斯网络分类器是数据挖掘与知识发现领域研究的主要方法之一.层次朴素贝叶斯分类器通过引入潜在节点来实现属性变量间存在聚集的层次关系,提出学习该分类器的构造算法.算法首先借助节点间的条件互信息值来锁定可能聚集节点的范围,然后再通过模拟退火算法来搜索评分较高的模型.层次朴素贝叶斯分类器的结构特点适于构造水质富营养化评价模型,应用于水质预警系统的结果证明该方法可行,并具有较好的适用效果.     

基于滑移辛几何模态分解的行星齿轮箱故障诊断研究

针对辛几何模态分解方法在分解复杂信号时的特征提取能力不足问题,提出了一种基于滑移辛几何模态分解(SSGMD)的故障诊断方法。首先,通过加窗的方式构造了滑移矩阵,以代替轨迹矩阵,增强了周期性特征提取能力;其次,对滑移矩阵进行了辛几何相似变换,获得了其特征值,将特征值所对应的特征向量经过重构,得到了其初始单分量矩阵;然后,对初始单分量矩阵做对角平均化,得到了一系列初始辛几何分量;最后,对这一系列初始辛几何分量进行拼接重组,得到了滑移辛几何分量(SSGCs),进而完成了对信号的自适应分解。研究结果表明：通过对仿真信号和行星齿轮箱实测信号进行实验分析,可知SSGMD利用滑移矩阵和辛几何相似变换不仅可以保护原始信号结构化信息不变,而且能充分提取原始信号的状态信息;与经典的信号分解方法相比,SSGMD方法能有效地对多分量信号进行分解,具有优越的特征提取能力。     

基于时频融合和注意力机制的深度学习行星齿轮箱故障诊断方法

针对行星齿轮箱振动信号频率成分复杂和时变性强的问题,提出了基于时频融合和注意力机制的深度学习行星齿轮箱故障诊断方法。首先,采用小波包分解将原始振动信号分解到频带和时间两个维度作为输入数据;然后,使用卷积神经网络融合数据的频带特征,使用双向门控循环单元融合时序特征;接着采用注意力结构对不同时间点的特征自适应地进行动态加权融合;最后通过分类器进行识别,实现行星齿轮箱的端对端故障诊断。实验表明,该方法对比现有的深度学习故障诊断模型具有更高准确率,能够对行星齿轮箱多种健康状态进行准确地诊断。     

基于全矢谱的风力发电系统齿轮箱的故障诊断

介绍了风力发电齿轮箱系统变载荷、低转速、工况复杂的运行特点,阐述了风力发电系统齿轮箱的主要故障特征,针对旋转机械单源振动信号信息具有不完善性的缺陷,提出了基于同源信息数据融合的思想,将全矢谱技术应用于风电齿轮箱的故障检测方法。结合北方某风电场19#风机齿轮箱故障检测的实例,将全矢谱技术应用于诊断实例且得到了正确的诊断结果,证实了将全矢谱技术应用于风电齿轮箱检测中是准确、有效、实用的旋转机械故障诊断方法。     

Fault level diagnosis for planetary gearboxes using hybrid kernel feature selection and kernel Fisher discriminant analysis

This paper introduces a hybrid dimension reduction method that combines kernel feature selection and kernel Fisher discriminant analysis (KFDA). In the first stage, a kernel feature selection method is proposed to remove redundant and irrelevant features for two purposes: (1) reducing computation burden of the entire fault diagnosis system and (2) alleviating the impact of irrelevant features on KFDA. In the second stage, KFDA is used to establish a more compact feature subset by extracting a smaller number of features. We use Gaussian radial basis function as the kernel function for the two kernel stages in the proposed method. A parameter selection method for this kernel is proposed to select the optimal values for the proposed method. Experimental results on fault level diagnosis demonstrate that the proposed hybrid dimension reduction method has advantages over other approaches that use feature selection or KFDA separately.     

Optimal design of a tilling machine reduction gearbox using Matlab

This paper describes the optimal design of the reduction gearbox of a tillage machine. The minimum center diameter was selected as the objective, and the contact fatigue strength, bending fatigue strength, condition of nonintervention, and oil film thickness ratio of the gearbox were applied as constraint conditions. The optimal model was solved by a Matlab program. The results show that the center diameter of the reduction gearbox decreased by about 10%. The resulting decrease in weight and volume led to a reduction in the amount of gearbox material and a consequent decrease in production cost.     

Mathematical simulation of epicycle gear vibrations in a planetary gearbox using finite element analysis

A model of forced vibrations in a planetary gearing based on the finite element and superelement method is proposed. This model is used to solve static and dynamic problems for mechanisms with planetary gearings.     

Fault diagnosis of spur gearbox based on random forest and wavelet packet decomposition

This paper addresses the development of a random forest classifier for the multi-class fault diagnosis in spur gearboxes. The vibration signal’s condition parameters are first extracted by applying the wavelet packet decomposition with multiple mother wavelets, and the coefficients’ energy content for terminal nodes is used as the input feature for the classification problem. Then, a study through the parameters’ space to find the best values for the number of trees and the number of random features is performed. In this way, the best set of mother wavelets for the application is identified and the best features are selected through the internal ranking of the random forest classifier. The results show that the proposed method reached 98.68% in classification accuracy, and high efficiency and robustness in the models.     

基于BP神经网络对减速箱齿轮副的故障诊断与仿真研究

齿轮故障诊断的基本方法是利用振动信号,通过提取特征信息,比较各次测量中的差异进行诊断。根据BP神经网络诊断方法,运用虚拟样机技术建立齿轮模型,模拟各种各种故障,提取振动信号,然后给出BP神经网络训练的特征频率,使其对数据进行分析学习。当输入待诊断样本后,BP网络对数据进行对比分析得出诊断结果。经过仿真检验表明该方法对减速箱齿轮副的故障诊断有效,对其他旋转机械的故障诊断和维修保养也具有一定的实用价值。     

基于朴素贝叶斯方法和权值分析方法的电机轴承故障诊断

为了分析电动机轴承故障类型,首先采用小波包分析对轴承振动信号进行了高低频分解及重构,然后以各频带的能量值构成了轴承振动信号的特征向量,最后通过朴素贝叶斯网和提出的权重分析两种方法进行了电机轴承故障分类;采用朴素贝叶斯网对已知电机轴承故障类型的样本数据进行训练,获得参数后识别未知样本的故障类型,利用权重分析法计算未知与已知类型的电机轴承振动样本的相关系数,然后构建权重,并按照权值和的大小获取未知样本的故障类型。仿真结果表明,朴素贝叶斯网能较好地实现电机故障诊断,所提出的权重分析方法也能较好地对电机故障进行诊断。     

Design and evaluation of two-stage planetary gearbox for special-purpose industrial machinery

Journal of Mechanical Science and Technology - The gearbox, as a traditional machine that changes the transmission ratio and transfers power safely, stably, quietly and efficiently, is an...     

基于贝叶斯网络的车身制造偏差诊断

车身结构的复杂性及知识表达的不精确性,使得车身故障症状与故障原因之间的映射表现为随机和不确定。针对这些特点,在大量车身测量数据和历史诊断案例的基础上,将贝叶斯网络引入到车身偏差故障诊断中去。对贝叶斯网络的参数学习进行了探讨,结合实例统计和相关性分析建立了车身偏差诊断的贝叶斯网络模型。最后用以某车型的偏差诊断案例对该方法进行了验证,结果表明该方珐在工程实际中有一定的指导性。     

Vibration mechanism and improved phenomenological model of the planetary gearbox with broken ring gear fault

Journal of Mechanical Science and Technology - Accurate modeling of the vibration signal model of planetary gearboxes is essential for the subsequent fault diagnosis. According to the existing...     

Analog fault diagnosis using S-transform preprocessor and a QNN classifier

A novel method for fault diagnosis in analog circuits using S-transform (ST) as a preprocessor and a quantum neural network (QNN) as a classifier is proposed in this paper. The ST provides a frequency-dependent resolution and the features obtained from ST are distinct, and easy to understand. The QNN identifier, a computational tool for fuzzy classification combining the advantages of neural modeling and fuzzy-theoretic principles, has the ability to autonomously detect the presence of uncertainty, adaptively learn the existing uncertainty, properly approximate any membership profile, and autonomously quantify uncertainty in sample data. The comparison between the ST-based method and the wavelet-transform-based method, and comparison between the QNN method and the traditional NN method for analog fault diagnosis is provided. Simulation results show that the proposed method is effective in enhancing the efficiency of the training phase and the performance of the fault diagnostic system. The results clearly indicate more than 97.61% correct classification of fault classes in the example circuits of various sizes in the presence of similar faults.     

Fault diagnosis of a gearbox using the sliced Wigner fourth order time frequency method smoothed by a new kernel function

The detection of impulsive signals embedded in the broadband noise is useful for the fault diagnosis of a gearbox. The sliced Wigner fourth-order time frequency method (SWFOTFM) has been used for the detection of impulsive signals embedded in the broadband noise. However, one disadvantage of SWFOTFM is that the non-oscillating cross-terms cannot be smoothed by conventional kernel functions. In this paper, a new kernel function is developed to reduce the non-oscillation cross-terms. The SWFOTFM using the new kernel function is successfully applied to the fault diagnosis of a gearbox.     

Rolling bearing fault diagnosis based on mean multigranulation decision-theoretic rough set and non-naive Bayesian classifier

To analyze data from multi-level view, reduce computational burden, and improve fault diagnosis accuracy, a novel fault diagnosis method of rolling bearings based on mean multigranulation decision-theoretic rough set (MMG-DTRS) and non-naive Bayesian classifier (NNBC) is proposed in this paper. First, fault diagnosis features of rolling bearings in training samples are extracted to construct MMG-DTRS. Then, the significance degree of condition attribute in MMG-DTRS is defined to quantitatively measure the influence of condition attributes with respect to the decision ability of an information system. An attribute reduction algorithm based on MMG-DTRS is applied to acquire a lower dimensional condition attribute set, which reduces computational complexity and avoids the interference of irrelevant or redundant condition attributes. Finally, NNBC is constructed to classify rolling bearing conditions in test samples. The classification procedures by using NNBC are given. The performance of the proposed method is validated and the advantages are investigated by using a fault diagnosis experiment of rolling bearings. Experimental investigations demonstrate the proposed method is effective and reliable in identifying fault categories and fault severities of rolling bearings.     

基于最小互信息准则的盲源分离在齿轮箱故障诊断中的应用

提出一种利用盲源分离技术对齿轮箱混合故障进行诊断的方法。该方法以最小互信息量为准则,采用自然梯度的自适应算法求解统计独立源信号的估计值,并根据分离信号的频谱成功地提取了混合故障的特征信息,有效地诊断出齿轮箱所处的故障状态。     

Fault diagnosis of a gearbox based on the analysis of a fractional energy gathering band

To extract the weak fault feature of the accelerating process from a gearbox, a fractional energy gathering band time–frequency aggregated spectrum (FETFAS) is proposed to achieve a fast time–frequency analysis of a large signal and to highlight target components. The best order of the fractional Fourier transform (FRFT) is determined according to the rotating speed signal and transmission ratio. The vibration signal from the accelerating process of a gearbox is processed using the best order FRFT. The energy gathering band (EGB) is determined from the modulus spectrum of the FRFT. Then, the result of the FRFT within the EGB is analyzed using time–frequency analysis, and the energy from this result is aggregated to form the FETFAS. The experimental results show that the method to determine the best order of the FRFT from the rotating speed signal is fast and accurate. The time–frequency analysis of the FRFT’s results in the EGB requires less computation and has a high resolution. The FETFAS has the ability to focus and zoom and is able to highlight the target components and restrain noise. Therefore, the FETFAS is an effective method to extract weak fault feature from the signal of gearbox’s accelerating process.     

行星变速箱机构性能和结构特征的评价指标分析

行星变速箱广泛应用于各种车辆的传动系中。为获得结构合理、紧凑、性能好、可靠的传动装置,组成行星变速箱的行星机构应有多个评价指标。主要研究这种机构的运动学指标、性能指标和结构特征指标,并剖析其内在关系,为设计时的合理选择提供理论和技术依据。