Application and improvement of wavelet packet de-noising in satellite transponder

The satellite transponder is a widely used module in satellite missions, and the most concerned issue is to reduce the noise of the transferred signal. Otherwise, the telemetry signal will be polluted by the noise contained in the transferred signal, and the additional power will be consumed. Therefore, a method based on wavelet packet de-noising （WPD） is introduced. Compared with other techniques, there are two features making WPD more suit- able to be applied to satellite transponders： one is the capability to deal with time-varying signals without any priori information of the input signals; the other is the capability to reduce the noise in band, even if the noise overlaps with signals in the frequency domain, which provides a great de-noising performance especially for wideband signals. Besides, an oscillation detector and an av- eraging filter are added to decrease the partial oscillation caused by the thresholding process of WPD. Simulation results show that the proposed algorithm can reduce more noises and make less distortions of the signals than other techniques. In addition, up to 12 dB additional power consumption can be reduced at -10 dB signal-to-noise ratio （SNR）.

Reduction of ultrasonic echo noise based on improved wavelet threshold de-noising algorithm for friction welding

In the ultrasonic detection of defects in friction welded joints, it is difficult to exactly detect some weak bonding defects because of the noise pollution. This paper proposed an improved threshold function based on the multi-resolution analysis wavelet threshold de-noising method which was put forward by Donoho and Johnstone, and applied this method in the de-noising of the defective signals. This threshold function overcomes the discontinuous shortcoming of the hard-threshold function and the disadvantage of soft threshold function which causes an invariable deviation between the estimated wavelet coeffwients and the decomposed wavelet coefficients. The improved threshold function is of simple expression and convenient for calculation. The actual test results of defect noise signal show that this improved method can get less mean square error （ MSE ） and higher signal-to-noise ratio of reconstructed signals than those calculated from hard threshold and soft threshold methods. The improved threshold function has excellent de-noising effect.

Application of Wavelet Packet De-noising in Time-Frequency Analysis of the Local Wave Method

The local wave method is a very good time-frequency method for nonstationaryvibration signal analysis. But the interfering noise has a big influence on the accuracy oftime-frequency analysis. The wavelet packet de-noising method can eliminate the interference ofnoise and improve the signal-noise-ratio. This paper uses the local wave method to decompose thede-noising signal and perform a time-frequency analysis. We can get better characteristics. Finally,an example of wavelet packet de-noising and a local wave time-frequency spectrum application ofdiesel engine surface vibration signal is put forward.

A real-time 5/3 lifting wavelet HD-video de-noising system based on FPGA

In accordance with the application requirements of high definition(HD) video surveillance systems,a real-time 5/3 lifting wavelet HD-video de-noising system is proposed with frame rate conversion(FRC) based on a field-programmable gate array(FPGA),which uses a 3-level pipeline paralleled 5/3 lifting wavelet transformation and reconstruction structure,as well as a fast BayesS hrink adaptive threshold filtering module.The proposed system demonstrates de-noising performance,while also balancing system resources and achieving real-time processing.The experiments show that the proposed system's maximum operating frequency(through logic synthesis and layout using Quartus 13.1 software) can reach 178 MHz,based on the Altera Company's Stratix III EP3SE80 series FPGA.The proposed system can also satisfy real-time de-noising requirements of 1920 × 1080 at60 fps HD-video sources,while also significantly improving the peak signal to noise rate of the denoising images.Compared with similar systems,the system has the advantages of high operating frequency,and the ability to support multiple source formats for real-time processing.

A continuous differentiable wavelet threshold function for speech enhancement

Enhanced speech based on the traditional wavelet threshold function had auditory oscillation distortion and the low signal-to-noise ratio (SNR). In order to solve these problems, a new continuous differentiable threshold function for speech enhancement was presented. Firstly, the function adopted narrow threshold areas, preserved the smaller signal speech, and improved the speech quality; secondly, based on the properties of the continuous differentiable and non-fixed deviation, each area function was attained gradually by using the method of mathematical derivation. It ensured that enhanced speech was continuous and smooth; it removed the auditory oscillation distortion; finally, combined with the Bark wavelet packets, it further improved human auditory perception. Experimental results show that the segmental SNR and PESQ (perceptual evaluation of speech quality) of the enhanced speech using this method increase effectively, compared with the existing speech enhancement algorithms based on wavelet threshold.

Denoising of an Image Using Discrete Stationary Wavelet Transform and Various Thresholding Techniques

Image denoising has remained a fundamental problem in the field of image processing. With Wavelet transforms, various algorithms for denoising in wavelet domain were introduced. Wavelets gave a superior performance in image denoising due to its properties such as multi-resolution. The problem of estimating an image that is corrupted by Additive White Gaussian Noise has been of interest for practical and theoretical reasons. Non-linear methods especially those based on wavelets have become popular due to its advantages over linear methods. Here I applied non-linear thresholding techniques in wavelet domain such as hard and soft thresholding, wavelet shrinkages such as Visu-shrink (non-adaptive) and SURE, Bayes and Normal Shrink (adaptive), using Discrete Stationary Wavelet Transform (DSWT) for different wavelets, at different levels, to denoise an image and determine the best one out of them. Performance of denoising algorithm is measured using quantitative performance measures such as Signal-to-Noise Ratio (SNR) and Mean Square Error (MSE) for various thresholding techniques.

Implementation of GPR Signals De-Noising Based on DSP

An important issue of ground-penetrating radar （GPR） signals analysis is de-noising thai is the guarantee of acquiring good detecting effect. The paper illustrates a successful application of digital single processor （DSP） based on wavelet shrinkage algorithm. In order to realize real-time GPP, signals analysis, some key issues are discussed such as the realization of fast wavelet transformation, the selection of CPU chip and the optimization of data movement. Experimenial results show that the DSP based application not only basically meets the real-time requirement of GPP, signals analysis, but also assures the quality of the GPR signals analysis.

联合小波-频域变换的自适应能量检测

针对传统能量检测方法在频谱感知领域中极易受低信噪比环境干扰,忽视可用频谱的定位亦会影响频谱状态的判别结果,提出了一种联合小波-频域变换的自适应能量检测方法,旨在提高能量检测的噪声灵敏度和判别精确度。通过离散小波包变换对信号进行分解并计算子带能量;结合能量范数降低自适应阈值的计算复杂度,以便与子带能量比较;采用快速傅里叶变换定位可用频谱范围。对该方法进行模拟仿真,探究自适应阈值与不同性能参数之间的变化关系。仿真结果表明,该方法具有良好的环境适配性与系统稳定性,且在不同信噪比环境下的检测误差更小。此外,对子带信号进行频域分析以实现归一化频率范围的重新排序,进一步提高了频谱感知的准确度。

基于增强多头注意力机制的Optuna-BiGRU测井岩性识别

测井岩性识别是油气勘探开发中至关重要的内容。针对现有算法模型在处理测井曲线数据时,无法有效捕获曲线内部深层关联和深度方向关系、拟合能力较弱、难以准确提取关键特征、噪声干扰以及模型超参数调优过程复杂困难等问题,提出了一种通过Optuna超参数优化双向门循环单元(Optuna-BiGRU)结合增强多头注意力机制(EMHA)的测井岩性识别模型——Optuna-BiGRU-EMHA模型。该模型引入残差机制和层归一化以改进多头注意力机制模块,并结合双向门循环单元(BiGRU)解决了处理测井数据时的问题,同时使用Optuna超参数优化框架和小波包自适应阈值方法分别解决了超参数调优和噪声干扰问题。首先通过交会图分析和敏感性箱线图分析选取自然伽马、深感应电阻率、中子-密度孔隙度、平均中子-密度孔隙度和岩性密度5个特征参数的测井数据,通过小波包自适应阈值方法对数据进行去噪,并将测井数据分割成数据块,然后利用Optuna框架优化BiGRU-EMHA模型超参数,最后通过实验对比K-近邻算法(KNN)、随机森林(RF)、极端梯度提升算法(XGBoost)、长短期记忆(LSTM)神经网络、BiGRU、双向长短期记忆(BiLSTM)神经网络、BiGRU-MHA、Optuna-BiGRU-EMHA等8种模型在测井岩性识别中的精度。结果表明:Optuna-BiGRU-EMHA模型识别准确率达到80%,相对于传统机器学习模型和深度学习模型,综合岩性识别准确率分别提高15.94%~23.14%和3.93%~15.94%,该模型为常规测井岩性识别提供了坚实的理论支持。

基于EEMD-WPT的温室环境数据优化处理研究

【目的】解决温室系统中的数据采集传感器容易受到多种环境因素的干扰,从而导致数据中存在噪声的问题。【方法】提出一种集合经验模态分解(Ensemble empirical mode decomposition,EEMD)与小波包自适应阈值(Wavelet packet adaptive threshold,WPT)算法联合的数据降噪处理方法,并采用卡尔曼滤波与自适应加权平均算法对降噪后的数据进行融合。【结果】将EEMD-WPT算法应用于含噪温、湿度数据的降噪处理,相较于降噪前的数据,信噪比提升了73.08%。该算法相较于传统WPT算法具有更好的降噪效果,处理后的数据信噪比提升了40.31%,均方根误差降低了84.75%。【结论】该算法能解决数据跳动、冗余和丢失等问题,并为温室控制系统提供了有效的参数,具有较大的实际应用价值。

参数优化VMD结合改进小波包阈值的去噪方法

针对轴承信号故障特征容易被噪声淹没的问题,提出一种参数优化变分模态分解结合改进小波包阈值的去噪方法。首先,通过变分模态分解(Variational Mode Decomposition,VMD)结合改进粒子群算法(Improve Particle Swarm Optimization,IPSO)将含噪信号分解为若干本征模态分量(Intrinsic Mode Function,IMF)。以最大相关系数-相关峭度为准则,把IMF分为高值分量(High-value Intrinsic Mode Function,HIMF)和低值分量(Low-value Intrinsic Mode Function,LIMF)。再对LIMF进行改进小波包(Improved Wavelet Packet,IWP)阈值去噪。最后对重构信号进行包络解调,提取轴承故障特征频率,完成故障诊断。实验结果表明,该方法不仅能够避免“过扼杀”现象,并且可以得到信噪比更高的去噪信号。

基于CEEMDAN和小波包分解的闸门振动信号降噪研究

针对闸门监测振动信号去噪问题,提出基于CEEMDAN(经验模态分解)和小波包分解的闸门振动信号降噪算法,通过采用CEEMDAN和小波包分解方法进行信号去噪,可以有效处理水电站泄洪闸门振动信号中受到的外部干扰。CEEMDAN方法能够将信号分解成多个本征模态函数(IMF),每个IMF代表不同频率的振动成分,使得外部干扰和真实信号成分可以分离。随后,小波包分解能够将每个IMF进一步分解成不同尺度和频率的子频带,这有助于更准确地定位和分离干扰成分。对每个子频带应用阈值去噪技术,可以有效去除噪声,保留真实信号。由测试结果可知,该算法能很好地剔除闸门振动信号中的无用噪声,有效提高闸门振动信号的准确性。

基于改进小波包能量熵和阈值自适应的切削颤振在线监测

颤振是影响机床加工质量的重要原因之一,传统的颤振监测算法对颤振孕育阶段的感知灵敏度低,且监测阈值的设定不具备泛化性和实时性,针对该问题提出了一种能够自适应地识别早期颤振的在线监测方法。首先使用改进的小波包能量熵算法(IWPEE)提取颤振特征,在提高识别精度和鲁棒性的同时降低了计算量。其次基于改进的拉依达准则确定颤振监测阈值,使系统能够根据不同的加工条件自适应地计算颤振监测阈值。然后根据实际加工监测需求开发高效颤振在线监测软件,并且通过仿真信号和切削试验验证了本文所提算法的有效性。结果表明,IWPEE算法相较于传统熵值判定法,识别灵敏度提高了360%,改进的拉依达准则能自适应地确定阈值并成功在颤振孕育阶段将其监测出来,相较于传统阈值算法在阈值稳定性和适应性上有显著提升。

基于RLMD与SSA-BP的采煤机工况模式识别

针对采煤机工作环境恶劣、振动噪声大、故障率高等特点,提出了一种基于鲁棒性局部均值分解(RLMD)结合麻雀搜索算法(SSA)优化BP神经网络的采煤机工况模式识别方法。首先,对不同工况下采集到的振动数据进行小波包阈值降噪,之后进行RLMD分解,选取相关系数大于0.1的P_(F)分量对信号进行重构,选取排列熵、样本熵、模糊熵作为每条数据的特征集,最后输入到SSA-BP神经网络中进行工况模式识别,并与PSO-BP、GA-BP的识别结果进行对比,仿真结果表明:该方法的迭代次数最少,同时准确率最高,可达98.33%;最后也对比了数据预处理前后的准确率,证实了所提及的小波包阈值降噪结合RLMD的双重降噪方法的有效性。

基于多频带多特征融合的配电网单相接地故障选段方法

针对现有配电网单相接地故障选段方法存在的通信负担重、阈值设置困难及故障特征量单一等问题,提出一种基于多频带多特征融合的配电网单相接地故障选段方法。首先,利用小波包变换分别对故障后首半个工频周期的零序电压导数和零序电流进行分解和重构,得到原始波形的不同频带分量。其次,根据不同频带分量间的极性关系分别引入伏安特性特征向量与零序功率累加和作为故障特征。最后,筛选特征频带并构造基于特征频带与故障特征融合的选段判据。仿真实验及现场数据验证结果表明,所提方法能够有效实现单相接地故障免阈值就地选段。

基于FastDTW的电力计量装置故障智能诊断技术

采用目前方法诊断电力计量装置故障时,没有去除电流信号噪声,存在F1-Score值低、电流信号清晰度低、诊断故障效率低的问题。为此,提出了基于FastDTW的电力计量装置故障智能诊断方法。结合EMD法和小波包能量法预处理电流信号,去除噪声,分析并排列高低频电流信号,重组去噪后的电流信号,获得清晰有效的电流信号。在此基础上采用扭曲电流路径距离法结合动态阈值,通过计算实现电力计量装置故障智能诊断。实验结果表明,所提方法的F1-Score值、电流信号清晰度及诊断故障效率均较高。

基于CEEMDAN-小波包自适应阈值混凝土声发射信号降噪研究

为了得到更加纯净的混凝土声发射(acoustic emission, AE)信号来更准确地监测混凝土结构破裂过程,提出了一种完全自适应噪声集合经验模态分解(complete ensemble empirical mode decomposition, CEEMDAN)与小波包自适应阈值联合方法对循环荷载作用下的混凝土声发射信号进行降噪处理,运用信噪比和快速傅里叶变化(fast Fourier transform, FFT)分析来验证所用方法的可行性。实验结果表明:结合CEEMDAN-小波包自适应阈值对混凝土声发射信号进行降噪的效果较好,能有效地保留混凝土声发射信号特征信息,对混凝土声发射信号降噪提供新的思路,为后续利用声发射信号分析混凝土结构内部微裂纹扩展及演化特征奠定基础。

小时间尺度网络中断故障识别数学建模仿真

中断故障识别是小时间尺度网络正常运行的关键技术,但故障识别过程易受信号强度、网络配置和服务器断开等干扰。为强化小时间尺度网络运行的稳定性,提出基于贝叶斯分类的小时间尺度网络中断故障识别数学建模方法。利用采集Agent模块获取小时间尺度网络信号,通过最大似然译码对信号完成校正处理,再根据矢量量化技术判断故障节点,完成故障区域的定位。基于此,采用小波包变换法提取故障区域特征,通过阈值修正模块判断故障类型,实现小时间尺度网络中断故障识别。仿真结果表明,所提模型的置信度高、识别效率高,具有更好的容错性能。

基于改进CEEMDAN的应答器上行链路信号降噪研究

针对在高速铁路复杂电磁环境中应答器上行链路(balise uplink,BU)信号传输受扰的问题,提出了一种基于自适应白噪声完备经验模态分解(complete ensemble empirical mode decomposition with adaptive noise,CEEMDAN)与小波包自适应阈值的联合降噪方法。首先,采用CEEMDAN算法将模拟BU信号分解为12个模态分量,根据相关系数判断分量为相关分量或无关分量;然后,相关分量经小波包降噪处理后重构为降噪后的BU信号;最后,选用信噪比(signal-noise ratio,SNR)和均方根误差(root mean square error,RMSE)作为评价指标,将该方法与目前广泛采用的6种降噪方法进行对比,信噪比提高了0.4861~6.144 dB,均方根误差降低了0.0549~11.091。为检验该方法的实际应用效果,采用联合降噪方法对实测BU信号进行降噪处理。仿真验证和实验验证的结果表明,采用联合降噪方法降噪后的BU信号不仅噪声分量得到了有效去除,而且信号特征保存完好,证明该方法能够应用于解决实际BU信号受扰问题。

Research and Application of New Threshold De-noising Algorithm for Monitoring Data Analysis in Nuclear Power Plant

Under the complex condition of nuclear power plant, all kinds of influence factors may cause distortion of on-line monitoring data. It is essential that on-line monitoring data should be de-noised in order to ensure the accuracy of diagnosis. Based on the research of wavelet analysis and threshold de-noising, a new threshold denoising method based on Mallat transform is proposed. This method adopts factor weighing method for threshold quantization. Through the specific case of nuclear power plant, it is verified that the algorithm is of validity and superiority.