The Research of Blind Source Separation (BSS) in Machinery Fault Diagnosis

Blind source separation (BSS) technology is very useful in many fields, such as communication, radar and so on. Because of the advantage of BSS that it can separate multi-sources even not knowing the mix-coefficient and the probability distribution, it can also be used in fault diagnosis. In this paper, we first use the BSS to deal with the sound from the machinery in fault diagnosis. We make a simulation of two sound sources and four sensors to test the result. Each source is a narrow-band source, which is composed of several sine waves. The result shows that the two sources can be well separated from the mixed signals. So we can draw a conclusion that BSS can improve the technology of sound fault diagnosis, especially in rotating machinery.

FREQUENCY OVERLAPPED SIGNAL IDENTIFICATION USING BLIND SOURCE SEPARATION

The concepts, principles and usages of principal component analysis （PCA） and independent component analysis （ICA） are interpreted. Then the algorithm and methodology of ICA-based blind source separation （BSS）, in which the pre-whitened based on PCA for observed signals is used, are researched. Aiming at the mixture signals, whose frequency components are overlapped by each other, a simulation of BSS to separate this type of mixture signals by using theory and approach of BSS has been done. The result shows that the BSS has some advantages what the traditional methodology of frequency analysis has not.

Blind-restoration-based blind separation method for permuted motion blurred images

A novel single-channel blind separation algorithm for permuted motion blurred images is proposed by using blind restoration in this paper. Both the motion direction and the length of the point spread function （PSF） are estimated by Radon transformation and extrema a detection. Using the estimated blur parameters, the permuted image is restored by performing the L-R blind restoration method. The permutation mixing matrices can be accurately estimated by classifying the ringing effect in the restored image, thereby the source images can be separated. Simulation results show a better separation efficiency for the permuted motion blurred image with various permutation operations. The proposed algorithm indicates a better performance on the robustness against Gaussian noise and lossy JPEG compression.

Blind source separation of ship-radiated noise based on generalized Gaussian model

When the distribution of the sources cannot be estimated accurately, the ICA algorithms failed to separate the mixtures blindly. The generalized Gaussian model （GGM） is presented in ICA algorithm since it can model non- Ganssian statistical structure of different source signals easily. By inferring only one parameter, a wide class of statistical distributions can be characterized. By using maximum likelihood （ML） approach and natural gradient descent, the learning rules of blind source separation （BSS） based on GGM are presented. The experiment of the ship-radiated noise demonstrates that the GGM can model the distributions of the ship-radiated noise and sea noise efficiently, and the learning rules based on GGM gives more successful separation results after comparing it with several conventional methods such as high order cumnlants and Gaussian mixture density function.

Single channel blind source separation based on ICA feature extraction

A new technique is proposed to solve the blind source separation (BSS) given only a single channel observation. The basis functions and the density of the coefficients of source signals learned by ICA are used as the prior knowledge. Based on the learned prior information the learning rules of single channel BSS are presented by maximizing the joint log likelihood of the mixed sources to obtain source signals from single observation, in which the posterior density of the given measurements is maximized. The experimental results exhibit a successful separation performance for mixtures of speech and music signals.

BLIND SIGNAL SEPARATION BASED ON ME AND STATISTICAL ESTIMATION

There are two major approaches for Blind Signal Separation (BSS) problem: Maximum Entropy (ME) and Minimum Mutual Information (MMI) algorithms. Based on the recursive architecture and the relationship between the ME and MMI algorithms, an Extended ME(EME) algorithm is proposed by using probability density function (pdf) estimation of the outputs to deduce the corresponding iterative formulas in BSS. Based on the simulation results, it can be concluded that the proposed algorithm has better performances than the traditional ME algorithm in convolute mixture BSS problems.

BLIND SPEECH SEPARATION FOR ROBOTS WITH INTELLIGENT HUMAN-MACHINE INTERACTION

Speech recognition rate will deteriorate greatly in human-machine interaction when the speaker's speech mixes with a bystander's voice. This paper proposes a time-frequency approach for Blind Source Seperation (BSS) for intelligent Human-Machine Interaction(HMI). Main idea of the algorithm is to simultaneously diagonalize the correlation matrix of the pre-whitened signals at different time delays for every frequency bins in time-frequency domain. The prososed method has two merits: (1) fast convergence speed; (2) high signal to interference ratio of the separated signals. Numerical evaluations are used to compare the performance of the proposed algorithm with two other deconvolution algorithms. An efficient algorithm to resolve permutation ambiguity is also proposed in this paper. The algorithm proposed saves more than 10% of computational time with properly selected parameters and achieves good performances for both simulated convolutive mixtures and real room recorded speeches.

About Multichannel Speech Signal Extraction and Separation Techniques

The extraction of a desired speech signal from a noisy environment has become a challenging issue. In the recent years, the scientific community has particularly focused on multichannel techniques which are dealt with in this review. In fact, this study tries to classify these multichannel techniques into three main ones: Beamforming, Independent Component Analysis (ICA) and Time Frequency (T-F) masking. This paper also highlights their advantages and drawbacks. However these previously mentioned techniques could not afford satisfactory results. This fact leads to the idea that a combination of those techniques, which is depicted along this study, may probably provide more efficient results. Indeed, giving the fact that those approaches are still be considered as being not totally efficient, has led us to review these mentioned above in the hope that further researches will provide this domain with suitable innovations.

A SIGNAL-ADAPTIVE ALGORITHM FOR BLIND SEPARATION OF SOURCES WITH MIXED KURTOSIS SIGNS

This paper addresses the problem of Blind Source Separation (BSS) and presents a new BSS algorithm with a Signal-Adaptive Activation (SAA) function (SAA-BSS). By taking the sum of absolute values of the normalized kurtoses as a contrast function, the obtained signal-adaptive activation function automatically satisfies the local stability and robustness conditions. The SAA-BSS exploits the natural gradient learning on the Stiefel manifold, and it is an equivariant algorithm with a moderate computational load. Computer simulations show that the SAA-BSS can perform blind separation of mixed sub-Gaussian and super-Gaussian signals and it works more efficiently than the existing algorithms in convergence speed and robustness against outliers.

BLIND SIGNAL SEPARATION OF LINEAR MIXTURE USING TRILINEAR DECOMPOSITION

This paper introduces a new source separation technique exploiting the time coherence of the source signals. The proposed approach relies only on stationary second order statistics. Blind Signal Separation (BSS) method using trilinear decomposition is proposed in this paper. Simulation results reveal that our proposed algorithm has the better blind signal separation performance than joint diagonalization method. Our proposed algorithm does not require whitening processing. Moreover, our proposed algorithm works well in the underdetermined condition, where the number of sources exceeds than the number of sensors.

A TIME-FREQUENCY BLIND SEPARATION METHOD FOR UNDERDETERMINED SPEECH MIXTURES

The proposed Blind Source Separation method(BSS),based on sparse representations,fuses time-frequency analysis and the clustering approach to separate underdetermined speech mixtures in the anechoic case regardless of the number of sources.The method remedies the insufficiency of the Degenerate Unmixing Estimation Technique(DUET) which assumes the number of sources a priori.In the proposed algorithm,the Short-Time Fourier Transform(STFT) is used to obtain the sparse rep-resentations,a clustering method called Unsupervised Robust C-Prototypes(URCP) which can ac-curately identify multiple clusters regardless of the number of them is adopted to replace the histo-gram-based technique in DUET,and the binary time-frequency masks are constructed to separate the mixtures.Experimental results indicate that the proposed method results in a substantial increase in the average Signal-to-Interference Ratio(SIR),and maintains good speech quality in the separation results.

Detection and Separation of Event-related Potentials from Multi-Artifacts Contaminated EEG by Means of Independent Component Analysis

Event-related potentials (ERP) is an important type of brain dynamics in human cognition research. However, ERP is often submerged by the spontaneous brain activity EEG, for its relatively tiny scale. Further more, the brain activities collected from scalp electrodes are often inevitably contaminated by several kinds of artifacts, such as blinks, eye movements, muscle noise and power line interference. A new approach to correct these disturbances is presented using independent component analysis (ICA). This technique can effectively detect and extract ERP components from the measured electrodes recordings even if they are heavily contaminated. The results compare favorably to those obtained by parametric modeling. Besides, auto-adaptive projection of decomposed results to ERP components was also given. Through experiments, ICA proves to be highly capable of ERP extraction and S/N ratio improving.

基于复杂度追踪的模态参数识别方法对比研究

复杂度追踪(complexity pursuit, CP)是求解振动信号盲源分离(blind source separation, BSS)问题的一类经典方法。用复杂度追踪估计解混矩阵主要有基于源信号复杂度计算的梯度下降(complexity pursuit-gradient descent, CP-GD)算法和基于时间可预测度的广义特征值分解(temporal predictability-generalized eigenvalue decomposition, TP-GED)算法。当前,这两种算法的关联性与算法性能尚缺乏研究,因此对这两种算法的等价性和计算性能进行了研究。首先,给出CP-GD和TP-GED两种算法的具体理论及算法流程;其次,利用二、三自由度振动系统直观地展示并对比解混向量对应的源信号复杂度及可预测度的变化规律;最后,通过对多工况下多自由度系统的模态参数识别算例,对比研究两种算法的精度及计算量。研究结果表明:在低阻尼比及高信噪比条件下,两种方法得到的解混矩阵是相同的;考虑到计算信号复杂度和梯度下降较为耗时,CP-GD算法计算代价要高于TP-GED算法。

基于FFT-MCC分析的ICA(BSS)盲不确定性消除

为了消除ICA(BSS)估计的幅值、相位及排序等盲不确定性,提出一种基于快速傅里叶变换与最大相关准则分析的ICA(BSS)估计源自适应校正方法。借助对原始传感观测及估计源的频谱分析,近似获得各本底源信号在观测信号中所占的比重———初始放大权值;基于最大相关准则优化调整ICA(BSS)估计源的相位,并对初始放大权值进行微调,从而消除ICA(BSS)估计的盲不确定性,实现源波形的恢复及其混合参数的估计。仿真试验结果证明了该方法的有效性,也表明它在复杂系统源识别或重建方面具有较大的应用潜力。

基于QR分解的类Jacobi联合对角化算法

为提高实矩阵集的近似联合对角化的盲源分离性能,避免平凡解,提出了一种基于QR分解的类Jacobi联合对角化算法.利用QR分解的数值稳定性,采用Jacobi旋转矩阵,将分离矩阵分解为多个初等三角矩阵和正交矩阵的乘积,利用Jacobi旋转矩阵的结构及矩阵变换后的相关元素求解最优参数,将高维矩阵最小化问题转化为一系列低维矩阵子问题,提升源信号恢复精度.通过求解简化的Frobenius范数目标函数降低算法复杂度.混合心电信号仿真结果表明,与QRJ2D,LUCJD,EGJLUD算法相比,本文算法在分离精度和收敛速度方面均有一定优势.

基于BSS的跳频通信抗部分频带噪声阻塞干扰方法

根据跳频信号与部分频带噪声阻塞干扰信号的近似统计独立性,提出一种基于盲源分离的跳频通信对抗部分频带噪声阻塞干扰方法。所提方法利用分离信号的二阶或高阶统计量构建目标函数引导分离矩阵迭代,实现跳频信号与部分频带噪声阻塞干扰信号的有效分离,从而提高跳频通信的抗干扰能力。仿真结果表明,所提方法可明显改善跳频通信在部分频带噪声阻塞干扰下的误码率性能,而且分离算法的处理时延很小,有望满足跳频通信的实际需求。

基于EMMD和BSS的单通道旋转机械故障诊断方法

针对在欠定的观测信号情况下,传统基于矩阵的盲源分离算法效果比较差的问题,提出一种基于极值域均值模式分解和盲源分离的单通道旋转机械信号故障特征提取方法,并应用于实际的故障诊断中。该方法先通过极值域均值模式分解法分解观测信号,把得到的固有模态函数和原观测信号一起组成新观测信号,从而实现了信号升维,使欠定问题转化为正定问题;然后,由奇异值分解和贝叶斯准则进行源数估计;最后,利用基于四阶累积量的特征矩阵联合对角化方法实现信号的盲分离。通过仿真,验证了该方法对旋转机械故障信号进行盲源分离的可行性。将提出的方法应用到齿轮和轴承系统的故障诊断中,进一步证明了该方法的有效性。

ASTFA-BSS方法及其在齿轮箱复合故障诊断中的应用

自适应最稀疏时频分析(adaptive and sparsest time-frequency analysis,ASTFA)方法以分解得到的单分量个数最少为优化目标,以单分量的瞬时频率具有物理意义为约束条件,使得到的分量更加合理;结合盲源分离,提出了一种基于ASTFA的盲源分离方法并应用于齿轮箱复合故障诊断中。该方法首先利用ASTFA将单通道源信号进行分解,然后利用占优特征值法进行源数估计,根据源数重组观测信号,最后对观测信号进行盲源分离得到源信号的估计。实验结果表明,该方法可以有效地对齿轮箱复合故障信号进行分离进而实现齿轮箱的复合故障诊断。