基于CEEMD和改进小波阈值的机械密封声发射信号降噪方法

声发射作为一种无损检测结束被广泛应用于多个领域,针对声发射信号难以从背景噪声种分离的问题,提出了一种基于互补集合经验模分解（CEEMD）与改进小波阈值相结合的降噪方法。首先对声发射信号进行CEEMD分解,通过峰宽占比确定信噪分量分界点,对噪声分量进行改进的小波阈值降噪,将降噪后的分量与其余分量进行信号重构得到最终降噪结果。通过仿真信号和机械密封声发射实验信号论证了本文方法相较于传统小波降噪和CEEMD强制降噪更有效。     

基于小波变换和压缩感知对煤岩体声发射信号降噪方法

针对煤岩体声发射信号具有随机非平稳、易受噪声信号干扰的特点,提出一种基于小波变换和压缩感知理论相结合的煤岩体声发射信号降噪方法.该方法利用噪声信号不具有稀疏性的特点,在小波分解的基础上进行压缩感知的压缩重构与降噪处理,使噪声信号与煤岩体声发射信号最大程度分离,该方法不仅达到降噪的目的,还避免了小波阈值滤波中阈值选取对降噪效果的影响.与小波阈值滤波、小波包降噪方法和压缩感知方法在选取不同压缩比、观测矩阵、重构算法对降噪效果进行对比.仿真结果表明,所提方法能够有效地抑制噪声信号的干扰,提高信号信噪比,降低均方误差,具有较强的实用价值.     

基于小波包分解与RVM的声发射信号识别

声发射信号识别目前已经成为检测大型储罐底板缺陷状况的普遍方法之一.针对现有声发射检测技术对原始信号的处理与分析方法的不足,对储罐底板腐蚀程度分类识别方法的不明确,基于小波包分解,对声发射传感器接收到的原始信号进行特征提取,得出了信号的特征向量,通过训练相关向量机,结合相关向量机模型对待测设备进行腐蚀类型归类识别,得出分析结论.应用该方法对大连某企业的储油罐进行实测数据检验,分析结论与储油罐实际状况相同,验证了该方法具有科学性与可行性,能够诊断大型储罐底板腐蚀状况的准确性和可靠性.     

强噪条件下基于小波降噪的陀螺仪声信号处理方法

为了将噪声诊断技术应用于陀螺仪性能分析,针对陀螺仪声音信号信噪比较低的特点,提出了一种小波降噪的新方法,通过尺度系数的比对有效地在复杂强噪声条件下提取出所需的陀螺仪音频信号,为性能分析提供了保证。实验验证降噪取得了较好的效果。     

煤岩声发射信号识别研究

针对煤岩破裂的声发射信号难以在复杂的噪声环境中识别的问题,提出了一种基于小波包分析和小波特征能谱系数分析的煤岩声发射信号识别方法。选取Symlets小波作为煤岩声发射信号分析的小波基函数,采用混合阈值算法对该信号进行去噪处理,提取出有用声发射信号,并采用Matlab软件分别对有用声发射信号和噪声信号的小波包分解进行仿真,得到两者的小波特征能谱系数和小波包特征向量。仿真结果表明,有用声发射信号特征向量的各级能量变化程度较大,噪声信号特征向量的能量变化较为稳定,从而可实现煤岩声发射信号的识别。     

基于DSP的声信号降噪系统的设计与实现

基于DSP TMS320VC5502和音频采样芯片TLV320AIC23构建了声信号降噪系统,分析了该系统的总体结构设计和软件处理流程设计.在DSP/BIOS控制机制下,建立了基于DMA、双缓冲区切换和中断响应机制等技术的音频信号采集与处理的程序框架.提出了一种基于多分辨分析的小波频带阈值降噪方法,并应用TDS5502EVM板和TDS560仿真器对优化的小波阈值降噪算法进行了实时仿真,实现了对强噪声语音信号的分析与降噪处理,并通过一个实例信号验证了系统的良好性能.     

不同充放电倍率下的锂电池声发射信号分析

锂离子电池充放电倍率是影响电池老化的主要外部应力之一,充放电倍率过大可能使电池内部电流分布不均,导致锂沉积或电极活性材料结构疲劳变形。基于声发射方法研究了不同充放电倍率下的电池状态。通过设计锂离子电池声发射信号试验平台,采集了不同充放电倍率下的电池声发射信号。对声发射信号进行参数与波形分析,发现声发射信号主要存在两个突发型的波形,且充电与放电的声发射信号波形的初次穿越阈值的采样点符号相反。锂离子电池声发射信号的幅度和波形时间间隔都与电池的充放电倍率相关,电池充放电倍率越大,声发射信号幅度越高,波形时间间隔越小。     

基于小波自适应算法的传感器信号降噪研究

基于MALLAT算法原理和自适应算法,设计了小波自适应算法的结构,并对算法进行了理论分析和仿真研究。仿真结果表明,小波自适应算法在传感器信号降噪方面表现出了良好的性能。     

乘性噪声下的弱GPS信号捕获

乘性噪声对弱GPS信号的检测与捕获具有重要影响。在分析乘性噪声的循环平稳特性的基础上,提出一种在乘性噪声条件下的弱GPS信号检测方法。首先通过本地产生的伪随机码与接收到的GPS信号进行位乘,以尝试性地消除伪码对GPS信号二阶统计量的周期性的影响;然后利用去伪码后的GPS信号的循环平稳特性估计GPS信号的载频,从而进一步地完成对GPS信号的检测。同时还提出了一种乘性噪声条件下的弱GPS信号检测结构,分析了基于该检测结构的检测性能。最后通过仿真验证了该方法的有效性。     

强噪声背景下钢丝绳损伤信号降噪方法

钢丝绳损伤信号是一种非平稳无周期性的冲击信号,其特征信号的降噪处理和特征提取成为亟待解决的难题。小波变换方法若小波基或者分解层数不适合,会在信号降噪的同时引入其他噪声干扰,影响信号处理与特征提取的效果。相较于小波变换方法,移位平均法只需要选择一定的移位窗宽即可实现对信号的有效降噪,但移位窗宽需要人为选择,盲目性大。针对上述问题,提出一种强噪声背景下钢丝绳损伤信号降噪方法。利用钢丝绳漏磁检测传感器采集不同类型的断丝数据,向信号中加入强高斯白噪声,以模拟强噪声背景;采用自适应移位平均法对钢丝绳损伤信号进行降噪,利用量子粒子群优化(QPSO)算法优化移位平均法的窗宽;将损伤信号的信噪比(SNR)作为适应度函数,通过QPSO算法使得损伤特征信号SNR最大化,从而实现最优信号降噪效果。实验结果表明,对于强噪声背景下的钢丝绳平稳和波动信号,相较于小波变换,自适应移位平均法的降噪效果更明显,信噪比更高,信号更为平滑。实测结果表明,对于现场采集的噪声相对弱一些的钢丝绳损伤信号,自适应移位平均法的降噪效果也比小波变换好,验证了自适应移位平均法具有较好的通用性。     

基于小波分解的磨床声发射特征信号提取方法

磨床磨削加工环境复杂,传统的信号分析方法难以对其进行有效的特征信号提取,提出了一种基于小波分解与能量谱相结合的特征信号提取方法.利用小波多尺度、多分辨率的特性,对磨床磨削的声发射信号进行多尺度分解;根据金属磨削的声发射信号的特性,选取最优频率段进行频谱分析,再结合能量谱提取特征信号.通过对自行研发的磨削监测系统工程试验分析表明:该方法能够有效、准确地诊断出磨削的每个状态,误判率为0.02％,相比单一的频谱分析诊断,精度更高、可靠性更好,具有一定的工程实用价值.     

A fast minimization method for blur and multiplicative noise removal

Multiplicative noise and blur removal problems have attracted much attention in recent years. In this paper, we propose an efficient minimization method to recover images from input blurred and multiplicative noisy images. In the proposed algorithm, we make use of the logarithm to transform blurring and multiplicative noise problems into additive image degradation problems, and then employ l ₁-norm to measure in the data-fitting term and the total variation to measure the regularization term. The alternating direction method of multipliers (ADMM) is used to solve the corresponding minimization problem. In order to guarantee the convergence of the ADMM algorithm, we approximate the associated nonconvex domain of the minimization problem by a convex domain. Experimental results are given to demonstrate that the proposed algorithm performs better than the other existing methods in terms of speed and peak signal noise ratio.     

MEMS加速度计噪声分析与降噪方法研究

针对微机电系统(MEMS)加速度计的随机噪声对输出信号干扰的情况,提出了对加速度计噪声源及噪声类型进行辨识、估计与建模,并确定误差补偿的降噪方法,以提高加速度计精度.采用Allan方差分析法对MEMS加速度计的随机噪声进行分析,得到了影响MEMS加速度计性能的几种主要随机噪声,使用自回归滑动平均模型(ARMA)对加速度计输出数据进行数学建模,以最终预测误差(FPE)准则确定使用的模型与阶次.设计了Kalman滤波算法,对加速度计进行降噪,通过Allan方差方法对Kalman算法滤波效果进行分析.实验结果表明:Kalman滤波能有效降低加速度计的随机噪声.     

Clustered and deep echo state networks for signal noise reduction

Machine Learning - Echo State Networks (ESNs) are Recurrent Neural Networks with fixed input and internal (hidden) weights, and adaptable output weights. The hidden part of an ESN can be considered...     

Improved subband-forward algorithm for acoustic noise reduction and speech quality enhancement

This paper addresses the problem of speech enhancement and acoustic noise reduction by adaptive filtering algorithms. Recently, we have proposed a new Forward blind source separation algorithm that enhances very noisy speech signals with a subband approach. In this paper, we propose a new variable subband step-sizes algorithm that allows improving the previous algorithm behaviour when the number of subband is selected high. This new proposed algorithm is based on recursive formulas to compute the new variable step-sizes of the cross-coupling filters by using the decorrelation criterion between the estimated sub-signals at each subband output. This new algorithm has shown an important improvement in the steady state and the mean square error values. Along this paper, we present the obtained simulation results by the proposed algorithm that confirm its superiority in comparison with its original version that employs fixed step-sizes of the cross-coupling adaptive filters and with another fullband algorithm.     

含噪声的海量多元时间序列降维方法研究

多元时间序列具有高噪声、非线性和海量的特点,但传统基于距离的降维方法难以有效的应对噪声带来的子空间偏移和数据的爆炸式增长。在基于角度优化的全局嵌入算法和共同核主成分分析方法的基础上,提出了一种基于角度优化的共同核主成分分析方法,并将该方法依托Hadoop平台进行了并行化改进,有效解决了噪音带来的子空间偏移和海量数据带来的巨大运算量问题。通过实验,对算法的有效性、运行效率及伸缩性进行了验证,结果表明提出的方法可以有效地对含有噪声的多元时间序列进行降维;基于Hadoop平台并行后的方法具有良好的运行效率和伸缩性。     

Acceleration of acoustic emission signal processing algorithms using CUDA standard

Offline processing of acoustic emission (AE) signal waveforms recorded during a long-term AE monitoring session is a challenging problem in AE testing area. This is due to the fact that today's AE systems can work with up to hundreds of channels and are able to process tens of thousands of AE events per second. The amount of data recorded during the session is very high.This paper proposes a way to accelerate signal processing methods for acoustic emission and to accelerate similarity calculation using the Graphic Processing Unit (GPU). GPU-based accelerators are an affordable High Performance Computing (HPC) solution which can be used in any industrial workstation or laptop. They are therefore suitable for onsite AE monitoring.Our implementation, which is based on Compute Unified Device Architecture (CUDA), proves that GPU is able to achieve 30 times faster processing speed than CPU for AE signal preprocessing. The similarity calculation is accelerated by up to 80 times. These results prove that GPU processing is a powerful and low-cost accelerator for AE signal processing algorithms.     

Parametrization method for calculating exact stability bounds of stochastic linear systems with multiplicative noise

For deterministic time-invariant linear systems, stability results are quite simple. For stochastic systems, however, even for linear ones, they are rather complicated. In this paper, some results on second mean stability (mean square stability) of time-invariant linear systems with multiplicative noise are summarized and the parametrization method of getting an exact bound for pth mean stability (p ≥ 2) via second mean stability is stated. Moreover, relations between pth mean stabilities for various values of p are given. On the basis of these relations, a simpler method for getting sufficient conditions for pth mean stability is also given, though the resulting sufficient bound is, of course, more conservative. Comparative studies of various conditions are made by using examples.