基于小波系数图和卷积神经网络的太赫兹光谱物质识别

许多太赫兹光谱物质识别方法依靠寻找该物质在太赫兹波段范围内不同光谱表现出的不同特征来识别特定物质。吸收峰提取法是常用的光谱特征提取算法,但当光谱无明显特征吸收峰或峰位、峰值相近或难以识别时,难以利用吸收峰特征辨别物质。将机器学习和统计学习技术用于太赫兹光谱的识别中虽减少了吸收峰的干扰,但常常需要人为定义特征而导致分类误差。深度学习法能自动提取特征,但在识别前往往需要进行复杂的预处理操作,并且在特征提取的过程中容易丢失部分特征从而导致分类误差。针对以上问题,提出了一种基于小波系数图和卷积神经网络的太赫兹光谱识别方法。利用太赫兹光谱信号进行小波变换时,由于小波系数矩阵的每一行系数与原始光谱信号存在着对应关系,因此将太赫兹光谱的吸收系数通过小波变换在频率域上展开,能得到不同的二维的频率-尺度分布图,又称小波系数图。然后构造一个卷积神经网络(CNN)对小波系数图进行分类,可得到太赫兹光谱物质的分类结果。为了验证所提出算法的有效性,将三组小波系数图数据与原始光谱数据分别输入CNN、Support Vector Machin (SVM)、Multilayer Perceptron (MLP)三种不同的分类器作对比,从实验结果可以发现本文算法在三组数据中的识别率均达到了100%,说明相比于传统方法,本文方法能准确分类没有明显特征吸收峰的光谱,证明了使用卷积神经网络识别小波系数图的有效性。为了体现本文算法的优势,与小波脊线寻峰识别算法作对比,实验结果表明本文算法几乎不受峰频、峰位、峰值的影响,无论是识别不存在吸收峰的淀粉,还是识别相似度高的蔗糖和葡萄糖,都具有较高的识别率,分类准确率达97.62%,证明了所提算法的优越性。该算法为太赫兹光谱数据识别提供了一种新思路,同时也可以推广运用到其他谱图物质的识别中。     

基于信息融合的模拟电路故障的特征提取与融合方法

在模拟电路故障诊断中,故障特征的提取是一个非常重要的环节,其提取结果的好坏将直接影响最终的诊断正确率。对现有文献研究发现,每种特征提取方法单独使用时都有一定的局限性,为了能够更加充分的提取模拟电路故障特征,提出了小波包分析与主元分析并行应用的方法,并将两种方法提取的特征向量依据不同规则进行了三种类型的融合,方便对比实验。为获取最优小波特征,提出了特征偏离度,并以此为标准选择最优小波基。最后,通过设计一种改进的神经网络分类器模型,将融合后的三种特征向量送入其中进行仿真验证,得出最终诊断结果。结果表明,该方法能够有效克服单一特征提取方法提取不充分的缺点,提高故障诊断的正确率,并且融合因子 适中时诊断正确率最高。     

A novel adaptive wavelet stripping algorithm for extracting the transients caused by bearing localized faults

Rolling element bearings are widely used in rotating machinery. Its unexpected failure may result in machine breakdown. Whenever a bearing suffers a localized fault, the transients with a potential cyclic characteristic are generated by the rollers striking the localized fault. This phenomenon is an early bearing fault feature. Therefore, the extraction of the transients is beneficial to the identification of the early bearing fault. In this paper, a novel adaptive wavelet stripping algorithm (AWSA) is proposed to extract the simulated transients from an original bearing fault signal. Firstly, the parametric model of anti-symmetric real Laplace wavelet (ARLW) or impulse response wavelet (IRW) is built to approximate the real transients. Then, with the aid of wavelet correlation filtering analysis, the simulated transients with the optimal frequency, damping coefficient and delay time are adaptively peeled from the original bearing fault signal. The spatial reconstruction of the simulated transients reflects the random occurrence of the real transients. In order to boost the computing time of the AWSA, an enhanced AWSA is developed. At last, the bearing fault signals collected from an experimental machine and an industrial machine are used to validate the effectiveness of the AWSA. The results show that the AWSA can adaptively peel the simulated transients from the original bearing fault signals. A comparison with a periodic multi-transient model is conducted to show that the AWSA is better to extract the random characteristics of the real transients.     

Identification of vehicle suspension shock absorber squeak and rattle noise based on wavelet packet transforms and a genetic algorithm-support vector machine

The squeak and rattle (S&R) noise of a vehicle’s suspension shock absorber substantially influences the psychological and physiological perception of passengers. In this paper, a state-of-the-art method, specifically, a genetic algorithm-optimized support vector machine (GA-SVM), which can select the most effective feature subsets and optimize the model’s free parameters, is proposed to identify this specific noise. A vehicular road test and a shock absorber rig test are conducted to investigate the relationship between these features, and then an approach for quantifying the shock absorber S&R noise is given. Pre-processed signals are decomposed through a wavelet packet transform (WPT), and two criteria, namely, the wavelet packet energy (WPE) and wavelet packet sample entropy (WPSE), are introduced as the feature extraction methods. Then, the two extracted feature sets are compared based on this genetic algorithm. Another advanced method, known as the genetic algorithm-optimized back propagation neural network (GA-BPNN), is introduced for comparison to illustrate the superiority of the newly developed GA-SVM model. The result shows that the WPSE can extract more useful features than the WPE and that the GA-SVM is more effective and efficient than the GA-BPNN. The proposed approach could be retrained and extended to address other fault identification problems.     

恒星大气物理参数估计中的稀疏特征提取

随着斯隆数字巡天项目(SDSS)、欧空局GAIA和我国大天区面积多目标光纤光谱天文望远镜(LAMOST)等项目的相继实施,拥有的恒星光谱数据量急速增加,由此导致基于光谱的恒星大气物理参数自动测量方法的研究成为天文光谱分析的重要课题之一[1]。探讨了恒星光谱特征提取的方法(Haar+lasso),其基本思想是首先使用Haar小波包对原始光谱进行多尺度分解,去除高频系数,选取低频系数作为光谱信息的描述;再采用lasso算法提取最优的特征; 最后将最优特征输入非参数回归模型中对恒星大气参数进行自动测量。Haar小波可以较好地去除原始光谱信号中的高频噪声,对全频谱数据进行降维。lasso算法可以进一步剔除数据冗余, 提取与物理参数相关度较强的特征。Haar+lasso方法提高了物理参数自动测量的准确性和运行效率。我们采用本文方案对SLOAN发布的40 000个恒星子样本的物理参数进行测量,三个物理参数的平均绝对误差为： log Teff: 0.007 1 dex,log g: 0.225 2 dex和[Fe/H]: 0.199 6 dex。同现有相关文献的实验结果相比,该方案可以获得更准确的物理参数。     

Wavelet filter-based weak signature detection method and its application on rolling element bearing prognostics

De-noising and extraction of the weak signature are crucial to fault prognostics in which case features are often very weak and masked by noise. The wavelet transform has been widely used in signal de-noising due to its extraordinary time-frequency representation capability. In this paper, the performance of wavelet decomposition-based de-noising and wavelet filter-based de-noising methods are compared based on signals from mechanical defects. The comparison result reveals that wavelet filter is more suitable and reliable to detect a weak signature of mechanical impulse-like defect signals, whereas the wavelet decomposition de-noising method can achieve satisfactory results on smooth signal detection. In order to select optimal parameters for the wavelet filter, a two-step optimization process is proposed. Minimal Shannon entropy is used to optimize the Morlet wavelet shape factor. A periodicity detection method based on singular value decomposition (SVD) is used to choose the appropriate scale for the wavelet transform. The signal de-noising results from both simulated signals and experimental data are presented and both support the proposed method.     

Low-Pass Filtering Empirical Wavelet Transform Machine Learning Based Fault Diagnosis for Combined Fault of Wind Turbines

Fault diagnosis of wind turbines is of great importance to reduce operating and maintenance costs of wind farms. At present, most wind turbine fault diagnosis methods are focused on single faults, and the methods for combined faults usually depend on inefficient manual analysis. Filling the gap, this paper proposes a low-pass filtering empirical wavelet transform (LPFEWT) machine learning based fault diagnosis method for combined fault of wind turbines, which can identify the fault type of wind turbines simply and efficiently without human experience and with low computation costs. In this method, low-pass filtering empirical wavelet transform is proposed to extract fault features from vibration signals, LPFEWT energies are selected to be the inputs of the fault diagnosis model, a grey wolf optimizer hyperparameter tuned support vector machine (SVM) is employed for fault diagnosis. The method is verified on a wind turbine test rig that can simulate shaft misalignment and broken gear tooth faulty conditions. Compared with other models, the proposed model has superiority for this classification problem.     

A united optimum images fusion based on analysis of color distortion

In remote sensing community, IHS (intensity, hue, and saturation) transform is one of the most commonly used fusion algorithm. A study on IHS fusion indicates that the color distortion cannot be avoided.Meanwhile, wavelet decomposition has a property of frequency division in transform domain. And the statistical property of wavelet coefficient reflects those significant features. So, a united optimal fusion method, which using the statistical property of wavelet decomposition and IHS transform on pixel and feature levels, is proposed. That is, the high frequency of intensity component I is fused on feature level with multi-resolution wavelet in IHS space, and the low frequency of intensity component I is fused on pixel level with optimal weight coefficients. Spectral information and spatial resolution are two performance indexes of optimal weight coefficients. Experiment results show that it is a practical and effective method.     

A new blind fault component separation algorithm for a single-channel mechanical signal mixture

A vibration signal collected from a complex machine consists of multiple vibration components, which are system responses excited by several sources. This paper reports a new blind component separation (BCS) method for extracting different mechanical fault features. By applying the proposed method, a single-channel mixed signal can be decomposed into two parts: the periodic and transient subsets. The periodic subset is related to the imbalance, misalignment and eccentricity of a machine. The transient subset refers to abnormal impulsive phenomena, such as those caused by localized bearing faults. The proposed method includes two individual strategies to deal with these different characteristics. The first extracts the sub-Gaussian periodic signal by minimizing the kurtosis of the equalized signals. The second detects the super-Gaussian transient signal by minimizing the smoothness index of the equalized signals. Here, the equalized signals are derived by an eigenvector algorithm that is a successful solution to the blind equalization problem. To reduce the computing time needed to select the equalizer length, a simple optimization method is introduced to minimize the kurtosis and smoothness index, respectively. Finally, simulated multiple-fault signals and a real multiple-fault signal collected from an industrial machine are used to validate the proposed method. The results show that the proposed method is able to effectively decompose the multiple-fault vibration mixture into periodic components and random non-stationary transient components. In addition, the equalizer length can be intelligently determined using the proposed method.     

联合稀疏表示在水下目标分类中的应用

水下目标分类识别的性能受所选特征的限制,多特征往往可以获得更加稳定的结果,针对这一问题,提出了一种基于联合稀疏表示模型的水下目标分类识别方法。首先对水下目标回波信号提取3种具有信息互补性与关联性的特征:中心矩特征、小波包能量谱特征、梅尔频率倒谱系数特征,然后应用加速近端梯度法对联合稀疏表示模型进行优化,求解得到最优联合稀疏系数,最后根据最小误差准则确定目标类别。在消声水池开展模拟实验,对6类目标进行分类识别,结果表明:与传统算法相比,提出的算法具有更高识别准确率,并且其执行效率较传统算法有很大提升。      

Envelope extraction based dimension reduction for independent component analysis in fault diagnosis of rolling element bearing

A robust feature extraction scheme for the rolling element bearing (REB) fault diagnosis is proposed by combining the envelope extraction and the independent component analysis (ICA). In the present approach, the envelope extraction is not only utilized to obtain the impulsive component corresponding to the faults from the REB, but also to reduce the dimension of vibration sources included in the sensor-picked signals. Consequently, the difficulty for applying the ICA algorithm under the conditions that the sensor number is limited and the source number is unknown can be successfully eliminated. Then, the ICA algorithm is employed to separate the envelopes according to the independence of vibration sources. Finally, the vibration features related to the REB faults can be separated from disturbances and clearly exposed by the envelope spectrum. Simulations and experimental tests are conducted to validate the proposed method.     

Deceptive Chinese speech detection based on sparse decomposition of cepstral feature

In order to improve the performance of deception detection based on Chinese speech signals, a method of sparse decomposition on spectral feature is proposed. First, the wavelet packet transform is applied to divide the speech signal into multiple sub-bands. Band cepstral features of wavelet packets are obtained by operating the discrete cosine transform on loga?rithmic energy of each sub-band. The cepstral feature is generated by combing Mel Frequency Cepstral Coefficient and Wavelet Packet Band Cepstral Coefficient. Second, K-singular value decomposition algorithm is employed to achieve the training of an over-complete mixture dictionary based on both the truth and deceptive feature sets, and an orthogonal matching pursuit algorithm is used for sparse coding according to the mixture dictionary to get sparse feature.Finally, recognition experiments axe performed with various classified modules. Experimental results show that the sparse decomposition method has better performance comparied with con?ventional dimension reduced methods. The recognition accuracy of the method proposed in this paper is 78.34%, which is higher than methods using other features, improving the recognition ability of deception detection system significantly.     

基于可见光的多波段偏振图像融合新算法

采用了一种新的基于小波变换的偏振图像融合算法.首先,将两个波段中的每一波段三幅偏振图像利用小波变换分解成低频和高频部分,低频的小波系数平均值作为融合后的低频系数,高频细节系数根据不同区域特征选择方法以及对应输入图像小波系数的窗口区域方差来确定融合后高频小波系数,得到一个波段一幅图像.接着,将得到的图像再进行小波分解,采用低频图像的小波系数最小值作为融合后的低频系数,高频图像根据纹理一致性测度的纹理检测确定融合规则,用来调整高频小波系数,将来自不同图像的特征与细节融合在一起,并对融合图像质量进行了对比评价.实验结果表明,融合后的偏振图像不仅反映了场景的偏振信息,而且还包含了丰富的光谱信息,目标与背景的衬比度也得到了增强,为进一步的目标检测和识别提供了便利.     

GUW-based structural damage detection using WPT statistical features and multiclass SVM

Recently, guided ultrasonic waves (GUW) are widely used for damage detection in structural health monitoring (SHM) of different engineering structures. In this study, an intelligent damage detection method is proposed to be used in SHM applications. At first, GUW signal is de-noised by discrete wavelet transform (DWT). After that, wavelet packet transform (WPT) is employed to decompose the de-noised signal and the statistical features of decomposed packets are extracted as damage-sensitive features. Finally, a multiclass support vector machine (SVM) classifier is used to detect the damage and estimate its severity. The proposed method is employed for GUW-based structural damage detection of a thick steel beam. The effects of different parameters on the sensitivity of the method are surveyed. Furthermore, by comparing with some other similar algorithms, the performance of the proposed method is verified. The experimental results demonstrate that the proposed method can appropriately detect a structural damage and estimate its severity.     

基于听觉谱特征的变压器绕组状态检测研究

绕组松动是变压器常见故障之一,对变压器的安全运行产生巨大威胁.故对其进行精准的监测,对提高电力系统的安全稳定性具有十分重要的意义.基于声信号的变压器绕组松动检测,由于其具有无损检测和不需停运变压器等优点,成为近年来研究的热点.但声信号检测存在故障特征提前复杂和易受噪声干扰等缺陷,限制了其工程应用.该文提出了一种基于声信...     

倒谱参数稀疏分解下的汉语音谎言检测

为了提高汉语语音的谎言检测准确率,提出了一种对信号倒谱参数进行稀疏分解的方法。首先,采用小波包滤波器组对语音信号进行多频带划分,求得子频带对数能量并进行离散余弦变换以提取小波包频带倒谱系数,结合梅尔频率谱系数得到倒谱参数;其次,依据K-奇异值分解方法分别利用说谎和非说谎两种状态下的语音倒谱参数集训练得到过完备混合字典,在此字典上根据正交匹配追踪算法对参数集进行稀疏编码提取稀疏特征;最终进行多种分类模型下的识别实验·实验结果表明,稀疏分解方法相比传统参数降维方法具有更好的优化性能,本文推荐的稀疏谱特征最佳识别率达到78.34%,优于其他特征参数,显著提高了谎言检测识别准确率。      

声纹的变压器放电与机械故障诊断研究

变压器中的一些放电和机械故障会产生异常声,可用于故障检测.据此,该文提出基于可听声的变压器放电和机械故障诊断方法.针对机械故障声与变压器本体噪声特征相似易混淆的问题,提出改进小波包-BP神经网络算法,与传统小波包-BP神经网络算法相比声音识别率提高了5.7％.为提高声音识别系统的泛化性,提出基于梅尔对数频谱和卷积神经网...     

空-谱二维蚁群组合优化SVM的高光谱图像分类

提出了一种空-谱二维特征蚁群组合优化支持向量机的高光谱图像分类算法。利用两类蚁群分别在光谱维空间和样本分布空间交替搜索最大类间距波段组合和异质样本,提取最优特征波段,降低了高光谱的波段信息冗余,去除训练样本中的异质样本,优化了训练样本特征空间分布。将蚁群组合优化后的高光谱图像和训练样本应用到支持向量机(SVM)分类器中,扩大了特征空间类间距,提高了SVM算法的分类精度。实验表明该算法总分类精度达95.45%,Kappa系数0.925 2,是一种分类精度较高的高光谱图像分类方法。     

新生儿疼痛面部表情识别方法的研究

针对新生儿的疼痛与非疼痛面部表情识别,提出将Gabor变换和支持向量机(SVM)相结合的分类识别方法.对归一化后的大小为112 pixel×92 pixel的新生儿面部图像进行二维Gabor小波变换,提取出412160维Gabor特征;针对Gabor特征向量维数高、冗余大的特点,采用Adaboost算法作为特征选择工具,去除冗余的Gabor特征,从412160维特征中选取出900维Gabor特征;对选取出的Gabor特征用SVM进行疼痛表情的分类识别.该方法综合运用Gabor特征对于面部表情的良好表征能力、AdaBoost算法的特征选择能力以及SVM在处理少样本、高维数问题中的优势.对510幅新生儿的表情图像进行测试的结果表明,疼痛与非疼痛表情的分类识别率达到85.29%,疼痛与安静表情的分类识别率达到94.24%,疼痛与哭表情的分类识别率达到78.24%.     

Chatter detection in band sawing based on discriminant analysis of sound features

In the paper chatter detection in band sawing is considered as a signal processing and classification problem. A multi-sensory experimental setup was established on an industrial band saw including sound, acceleration and cutting force, and measurements. Based on an experimental analysis sound signal is shown to be the most appropriate for chatter detection, therefore a sound-based online chatter detection method is proposed. The method consists of a sound signal pre-processing with Short-Time Fourier Transform, extraction of features in frequency space with optimal threshold and application of Quadratic Discriminant Analysis for chatter detection. The proposed method tested with twofold cross validation yields over 96% success of chatter detection.