oalib
Search Results: 1 - 10 of 100 matches for " "
All listed articles are free for downloading (OA Articles)
Page 1 /100
Display every page Item
稀疏自编码深度神经网络及其在滚动轴承故障诊断中的应用
Application of Deep Neural Network with Sparse Auto-encoder in Rolling Bearing Fault Diagnosis
 [PDF]

汤芳,刘义伦,龙慧
- , 2018,
Abstract: 针对目前滚动轴承故障诊断主要采用监督式学习提取故障特征的现状,提出了一种基于稀疏自编码的深度神经网络,实现非监督学习自动提取滚动轴承振动信号的内在特征用于滚动轴承故障诊断。首先,将轴承故障振动信号的频谱训练稀疏自编码获得参数;然后用稀疏自编码获得的参数和轴承振动信号频谱的频谱训练深度神经网络,并结合反向传播算法对深度神经网络进行整体微调提高分类准确度;最后用训练好的深度神经网络来识别滚动轴承故障。对正常轴承、外圈点蚀故障、内圈点蚀故障和滚动体裂纹故障振动信号的分析结果表明:相比反向传播神经网络,提出的深度神经网络更能准确的识别滚动轴承故障类型。
To overcome the problem of using supervised learning to extract fault features for most current rolling bearing fault diagnosis methods, a deep neural network algorithm is proposed, which is realized sparse auto-encoder, to achieve unsupervised feature learning by automatic extracting the inherent characteristics of the rolling bearing vibration signal for fault diagnosis of rolling bearing fault diagnosis. Firstly, the spectrum of the bearing vibration signal is used to train sparse auto-encoder in order to obtain parameters; secondly, the parameters from sparse auto-encoder and spectrum of the rolling bearing vibration signal are used to train the deep neural network, and the back-propagation algorithm is used for fine-tuning the deep neural network with the purpose of improving classification accuracy. Finally, the deep neural network has been trained to identify faults of rolling bearings. The analysis results from vibration signals with roller normal condition of the rolling bearing,pitting fault of bearing outer ring, pitting fault of bearing inner ring and crack fault of bearing rolling element show that, compared with back propagation neural network, the proposed deep neural network can accurately identify fault type of rolling bearing faults
堆叠自编码网络性能优化及其 在滚动轴承故障诊断中的应用
Optimization of Stacking Auto??Encoder with Applications in Bearing Fault Diagnosis
 [PDF]

张西宁,向宙,夏心锐,李立帆
- , 2018, DOI: 10.7652/xjtuxb201810007
Abstract: 为了解决堆叠自编码网络在参数较多时的梯度弥散问题,对网络每层的编码值进行了统计分析,发现大部分分布于激活函数的饱和区,这直接导致了神经元权值梯度的消失。为此,引入了一种标准化策略,将神经元按照样本进行归一化,然后引入两个待学习参数进行缩放和平移,最后通过激活函数输出到下一级神经元。运用带标准化的堆叠自编码网络进行滚动轴承故障诊断,将振动信号的频谱输入到网络中。与普通堆叠自编码网络相比,该标准化策略可有效地使网络编码值均匀分布,如将第一层编码值的熵从0.88 bit提高到了16.29 bit。带标准化的堆叠自编码网络可有效提高网络的抗噪能力和训练速度:在凯斯西储大学滚动轴承数据集上,当人为添加噪声信号的信噪比为0 dB时,识别正确率从16.18%提高到了100%;在实验室实测数据集上,不仅训练时间下降了37.22%,而且识别正确率从97.93%提高到了99.95%。对网络的编码值进行分析以及引入的标准化策略,可为科研技术人员构建堆叠自编码网络时提供参考,也为滚动轴承故障诊断提供了一种策略。
To solve the problem of gradient dispersion in stacking auto??encoder (SAE) with large number of parameters, we analyze the distribution of encoding values in each hidden??layer of network. It is found that most of them are distributed in the saturation area of the activation function, which directly leads to a weight gradient loss, thus a normalizing strategy is introduced. The node is normalized according to the sample, then two parameters are introduced to scale and move the encoding values. Then the modified values are passed to the activation function to next layer. The fault diagnosis of rolling bearing is carried out by the normalized SAE, and the spectrum of vibration signal is input into the network. Compared with ordinary SAE, the encoding values of normalized SAE are more well??distributed. For example, the entropy of encoding values in the first level is increased from 0.88 bit to 16.29 bit. The normalized SAE has higher anti??noise ability and faster training rate. When the signal to noise ratio (SNR) is 0 dB, the recognition accuracy is increased from 16.18% to 100% on the rolling bearing data sets of Case Western Reserve University. On laboratory data sets, the training time is decreased by 37.22%, and the recognition accuracy is increased from 97.93% to 99.95%. The introduced normalizing strategy provides a reference for subsequent research on the construction of SAE, and also provides a strategy for fault diagnosis of rolling bearings
基于自动编码器和SVM的轴承故障诊断方法
The Application of SVM Based on Auto-encoder in Bearing Fault Diagnosis
 [PDF]

雷文平,吴小龙,陈超宇,林辉翼
- , 2018,
Abstract: 支持向量机(Support Vector Machine, SVM)应用于轴承故障诊断前,首先要提取轴承的特征信号。在以往的特征信号提取中,往往是依据已有的知识模型进行特征筛选。随着近年来深度神经网络(Deep Neural Network, DNN)的应用与推广,自动编码器(Auto-encoder, AE)在特征提取方面的优势尤为突出。作为一种无监督的学习方式,AE能够基于数据驱动地提取信号的特征值,使得特征提取不再依赖于先验知识,从而让整个故障诊断过程更具智能化。本文运用改进的AE、去噪自动编码器(denoising autoencoder,DAE),进行轴承信号特征提取,并用SVM进行故障诊断。最终与基于经验模态分解(empirical mode decomposition, EMD)能量熵的SVM对比,反应具有无监督学习方式的DAE-SVM在轴承故障诊断方面的优越性,诊断准确率接近100%。
The fault feature should be extracted before the SVM was applied to the bearing fault diagnosis. In the previous feature signal extraction, it was often based on the existing knowledge model. With the application and promotion of DNN in recent years, AE had a special advantage in feature extraction. As an unsupervised learning method, AE could extract the features of the signal based on data driven, making the feature extraction no longer depends on prior knowledge, and the whole fault diagnosis processed more intelligent. In this paper, the improved AE、DAE,were used to extract the features of the bearing signals, and the fault diagnosis was carried out by SVM. Finally, by compared with the SVM based on EMD energy entropy feature extraction, the superiority of DAE-SVM with unsupervised learning method was reflected in beraing fault diagnosis, and its diagnostic accuracy was neraly 100%
滚动轴承故障检测深度卷积稀疏自动编码器建模研究
Rolling Bearing Fault Detection using Deep Convolution Automatic Sparse Encoder
 [PDF]

冯玉伯,丁承君,陈雪
- , 2018,
Abstract: 针对机械设备故障诊断大多采用有监督学习提取故障特征,而有标签数据难以获取的现状,提出一种在稀疏自动编码器中嵌入卷积网络的深度神经网络。利用希尔伯特和傅里叶变换实现机械设备振动时间序列向Hilbert包络谱的转换,通过卷积网络中多组卷积核自动学习谱空间数据的不同特征,保证了特征提取的自动化、全面性和多样性,稀疏自动编码器搜索具有正交性数据特征的低维表示,并使得编码后的数据具有很强的聚类特性,实现设备的自动故障诊断。通过对滚动轴承振动信号进行分析实验,证明该方法在设备故障诊断中具有去标签化、自动化、鲁棒性等特点。
The supervised learning is commonly adopted in fault feature extraction for mechanical equipment fault diagnosis, while the labeled data are often hard to obtain. To deal with such problem, a deep neural network embedding convolution networks in sparse encoder is proposed. The transformations of Hilbert and Fourier make it possible to transform the vibration time series of machinery into Hilbert envelope spectrum. Different features of spectral space data are automatically learned with multiple sets of convolution kernels in convolution networks, which ensure the automation, comprehensiveness and diversity of the extracted features. The sparse encoder looks for a low-dimensional representation of data featured with orthogonality, making the encoded data characterized by strong clustering; so that the automatic fault diagnosis of the equipment is realized. Through analysis and experiments on vibration signals of rolling bearings, it is proved that this method has the characteristics of de-labeling, automation and robustness in equipment fault diagnosis
滚动轴承故障特征提取的EMD频谱自相关方法
Feature Extracting Method in the Rolling Element Bearing Fault Diagnosis Based on EMD and Spectrum Auto-correlation
 [PDF]

万书亭,詹长庚,豆龙江
- , 2016, DOI: 10.16450/j.cnki.issn.1004-6801.2016.06.020
Abstract: 首先,在论述频谱自相关方法(spectrum auto-correlation,简称SAC)的特点、经验模态分解(empirical mode decomposition,简称EMD)分析过程和轴承故障机理的基础上,指出了在故障信号不占主导作用时频谱自相关方法在轴承故障诊断中的局限性,并得到仿真算例验证;然后,提出了基于经验模态分解和频谱自相关的轴承故障特征提取方法,将经验模态分解得到的各分量进行分析比较,再对适合的分量进行频谱自相关分析,可有效提出轴承故障频率;最后,分别在轴承故障试验台实测了深沟球轴承和圆柱滚子轴承内外圈故障振动数据,结果表明,EMD频谱自相关分析方法可以很好地提取轴承故障信号,较单一EMD分解、频谱自相关和峭度等方法效果更好,为轴承故障诊断提供了新思路。
This paper pointed out and verified the limitations of the spectrum auto-correlation (SAC) method based on the analysis process of empirical mode decomposition (EMD) and mechanism of bearing fault. Then, a feature extracting method of bearing fault diagnosis combined with EMD and SAC (EMD-SAC) was proposed, which analyzed the intrinsic mode function (IMF) based on EMD and chose the appropriate component to analyze using SAC. With this method, the bearing failure frequency could be separated from the complex signal. Finally, the inner and outer ring fault of deep groove ball bearings and cylindrical roller bearings respectively were tested in laboratory. The fault vibration data was measured and analyzed to verify the theoretical analysis presented above.
Fault Diagnosis of Rolling Bearing Based on Fast Nonlocal Means and Envelop Spectrum  [PDF]
Yong Lv,Qinglin Zhu,Rui Yuan
Sensors , 2015, DOI: 10.3390/s150101182
Abstract: The nonlocal means (NL-Means) method that has been widely used in the field of image processing in recent years effectively overcomes the limitations of the neighborhood filter and eliminates the artifact and edge problems caused by the traditional image denoising methods. Although NL-Means is very popular in the field of 2D image signal processing, it has not received enough attention in the field of 1D signal processing. This paper proposes a novel approach that diagnoses the fault of a rolling bearing based on fast NL-Means and the envelop spectrum. The parameters of the rolling bearing signals are optimized in the proposed method, which is the key contribution of this paper. This approach is applied to the fault diagnosis of rolling bearing, and the results have shown the efficiency at detecting roller bearing failures.
Resonance-Based Nonlinear Demodulation Analysis Method of Rolling Bearing Fault  [PDF]
Lingli Cui,Daiyi Mo,Huaqing Wang,Peng Chen
Advances in Mechanical Engineering , 2013, DOI: 10.1155/2013/420694
Abstract: Numerous mechanical nonstationary fault signals are a mixture of sustained oscillations and nonoscillatory transients, which are difficult to efficiently analyze using linear methods. We propose a nonlinear demodulation analysis method based on resonance and apply it to the fault diagnosis of rolling bearings. Unlike conventional demodulation methods that use frequency-based analysis and filtering techniques, our nonlinear demodulation analysis method is a decomposition demodulation of the signals according to different resonance based on Q-factors. When a local rolling bearing fault such as pitting is present, the fault vibration signals consist of the regular vibration signals and noise (a high resonance component containing multiple simultaneous sustained oscillations) and a transient impulse signal (a low resonance component being a signal containing nonoscillatory transients of faults). The regular vibration signal is a narrowband signal that has a high Q-factor, and the transient impulse signal is a wideband signal that has a low Q-factor. Using our resonance-based nonlinear demodulation analysis method, we decompose the signal into high resonance, low resonance, and residual components. Then, we perform a demodulation analysis on the low resonance component that includes the fault information. We have verified the feasibility and validity of the algorithm by analyzing the results of experimental and engineering signals. 1. Introduction The rolling bearing is one of the most widely used general mechanical components in rotating machines. Its running state directly affects the performance of the whole machine. Detecting and diagnosing its faults can prolong service life and reduce production costs. Therefore, it is important to monitor the bearing’s condition to ensure operational safety, prevent serious accidents, and reduce production costs [1]. The local fault vibration signal of a rolling bearing is a typical nonstationary signal. Compared with a stationary signal, the distribution parameters and regularities of the rolling bearing fault vibration signal are time dependent. Moreover, massive noise can be inevitably introduced into its vibration signal during the generation and transmission process, because of complex operational conditions and harsh environments. This is a major inconvenience to the analysis, handling, and use of the signals. When the inner ring, outer ring, or the rolling element of the rolling bearing is damaged, a periodic mechanical impact occurs in the contact between the surface of the fault and the surfaces of other
Multiscale Permutation Entropy Based Rolling Bearing Fault Diagnosis  [PDF]
Jinde Zheng,Junsheng Cheng,Yu Yang
Shock and Vibration , 2014, DOI: 10.1155/2014/154291
Abstract: A new rolling bearing fault diagnosis approach based on multiscale permutation entropy (MPE), Laplacian score (LS), and support vector machines (SVMs) is proposed in this paper. Permutation entropy (PE) was recently proposed and defined to measure the randomicity and detect dynamical changes of time series. However, for the complexity of mechanical systems, the randomicity and dynamic changes of the vibration signal will exist in different scales. Thus, the definition of MPE is introduced and employed to extract the nonlinear fault characteristics from the bearing vibration signal in different scales. Besides, the SVM is utilized to accomplish the fault feature classification to fulfill diagnostic procedure automatically. Meanwhile, in order to avoid a high dimension of features, the Laplacian score (LS) is used to refine the feature vector by ranking the features according to their importance and correlations with the main fault information. Finally, the rolling bearing fault diagnosis method based on MPE, LS, and SVM is proposed and applied to the experimental data. The experimental data analysis results indicate that the proposed method could identify the fault categories effectively. 1. Introduction The vibration signals of mechanical systems, especially for ones with fault, often show mutation, nonlinearity, and nonstationarity because of the strike, velocity chopping, structure transmutation, loading, and friction. Hence, it is very crucial for mechanical fault diagnosis to extract the fault feature information from the nonlinear and nonstationary signal. A primary method for dealing with the nonlinear and nonstationary signal is time-frequency analysis [1], which has been applied to the mechanical fault diagnosis field widely for its ability to provide local information both in time and frequency domains of vibration signals [2]. However, the time-frequency analysis method, such as wavelet transform or Hilbert-Huang transform [3, 4], which decomposes the vibration signal into several stationary monocomponent signals, cannot reflect the subtle dynamic changes of vibration signal effectively and, therefore, inevitably will have some limitations [5]. With the development of nonlinear dynamic theories, especially in recent years, a number of nonlinear parameters and methods, such as chaos theory, fractal dimension, and information entropy, have been applied to machine condition monitoring and fault diagnosis. For instance, Logan and Mathew elaborated the application of the correlation dimension to vibration fault diagnosis of rolling element bearing
A Fault Detection Method of Rolling Bearing Based on Wavelet Packet-cepstrum  [cached]
Jun Ma,Jiande Wu,Xiaodong Wang,Yugang Fan
Research Journal of Applied Sciences, Engineering and Technology , 2013,
Abstract: In this study, we put forward a fault detection method of rolling bearing based on the wavelet packet-cepstrum. Firstly, the original signal is decomposed using the wavelet packet. Secondly, calculate the energy of the decomposed sub-band reconstruction signal and select the relatively band which is concentrated on the fault energy. Finally, calculate cepstrum of the reconstruction signal to detect fault. The actual normal and fault data of the rolling bearing's outer ring is analyzed in applying this method in the MATLAB simulation circumstance. The result shows that the outer ring's failure frequency measured by the experiment is consistent with the theoretical calculation result.
一种新型深度自编码网络的滚动轴承健康评估方法
Deep auto-encoder network method for health assessment of rolling bearings
 [PDF]

佘道明,贾民平,,
- , 2018, DOI: 10.3969/j.issn.1001-0505.2018.05.004
Abstract: 为了准确描述滚动轴承性能退化的动态过程,结合深度学习强大特征提取能力的优势,提出了一种新型深度自编码和最小量化误差方法相结合的滚动轴承全寿命健康评估方法.用深度自编码模型对原始特征进行压缩提取,将压缩特征按趋势进行排序,选取趋势大的特征运用最小量化误差方法构建健康指标.针对基于一个度量的评价准则常具有偏差的问题,提出基于遗传算法的融合评价准则.2组实例分析结果表明,用该方法构建的健康指标的趋势值、单调性值、鲁棒性值、融合评价准则值都大于单层的自编码模型(AE)和传统的PCA降维方法,第1个实例中,该方法构建的健康指标融合评价准则值比PCA,AE方法分别增加了13.30%,3.17%;第2个实例中,该方法构建的健康指标融合评价准则值比PCA,AE方法分别增加了9.68%,3.85%.基于遗传算法的融合评价准则比单一的评价准则更具有说服力.
To describe the dynamic process of rolling bearing performance degradation accurately, considering the advantage of strong feature extraction ability of deep learning, a novel method combining deep auto-encoder(DAE)with minimum quantization error(MQE)was proposed to evaluate the whole life health of rolling bearings. The original feature was compressed and extracted by the DAE model, and the compressed feature was sorted according to the trend. Then, the feature with large trend was selected to construct the health index by using the MQE method. The evaluation criterion based on a single metric was often biased, so a fused evaluation criterion based on genetic algorithm was proposed. The superiority of the proposed method was demonstrated by comparison with the single-layer auto-encoder(AE)and principal components analysis(PCA)method. Two groups of examples show that the health index constructed by the proposed method is superior to other two methods in four aspects: trendability, monotonicity, robustness, and values of the fused criterion. In the first example, the values of the fused criterion of the health index constructed by the proposed method are 13.30% and 3.17% higher than those of PCA and AE method, respectively. In the second example, the value of fused criterion of health index constructed by the proposed method are 9.68% and 3.85% higher than those of PCA and AE method, respectively. The fused evaluation criterion based on the genetic algorithm is more persuasive than the single evaluation criterion
Page 1 /100
Display every page Item


Home
Copyright © 2008-2017 Open Access Library. All rights reserved.