自适应噪声抵消技术的应用研究与仿真

从噪声中分离有用信号是信号处理领域一个非常重要的课题.传统的滤波方法对于淹没于强干扰噪声背景下的有用信号提取已经不适用,而自适应滤波技术以其原理清楚,可实现性强而具有很强的应用价值.随着自适应技术的发展,自适应本舰噪声抵消技术已成为舰载声纳的信号处理技术中重要的研究课题之一.通过对自适应噪声抵消器原理的研究,结合舰载声纳信噪模型分析,重点研究利用自适应噪声抵消技术从舰载声呐在本舰强噪声干扰背景下提取有用信号的技术,并运用MATLAB进行了仿真试验.仿真结果表明自适应噪声抵消技术可以有效抵消本舰发出的强噪声,从而抑制本舰噪声对舰载声呐的干扰.     

自适应噪声抵消技术的仿真与应用研究

从有用信号中剔除噪声是信号处理中一个重要研究课题.在金属矿区进行地震勘探数据采集时,经常会受到矿区内机械设备的强噪声干扰,常规的滤波方法对于淹没在强机器噪声下的有效地震信号的提取已经不适用,而随着计算机和信号处理技术的发展,自适应噪声抵消技术已广泛的应用于各个领域.通过对自适应噪声抵消器原理的研究,结合地震数据信噪模型分析,重点研究利用自适应噪声抵消技术来消除机器噪声.针对基于NLMs算法的常规技术存在的不足,提出了一种改进的算法,并运用MATLAB进行仿真试验,仿真结果表明自适应噪声抵消技术可以有效抵消机器噪声的干扰,大大提高地震资料的信噪比,使有效地震信号失真小.     

基于神经元网络的自适应噪声抵消系统研究与实现

自适应噪声抵消技术是自适应滤波器最普遍的应用之一,它是一种在未知信号和噪声的先验知识条件下,能够很好地消除背景噪声影响的信号处理技术,具有很高的应用价值;但是,在很多情况下,噪声环境非常复杂,往往是非线性的,而目前所使用的自适应滤波器均属线性滤波器,滤波后会使原始信号产生失真;由于神经网络具有非线性等优点,可以很好的逼近非线性函数,所以提出了基于神经网络的自适应噪声抵消器;仿真结果表明,该方法可以效地实现噪声的抵消;最后提出应用DSP实现语音信号自适应噪声抵消的具体方案。     

基于神经元网络的自适应噪声抵消系统研究

在信号处理的研究中,自适应噪声抵消技术广泛地应用于通信、控制等领域,LMS是最常用的自适应算法,若信号通道结构比较复杂或存在非线性时,自适应滤波器的长度会增加,造成稳态失调、收敛速度降低,影响系统的性能.由于神经网络经过训练后可以很好地逼近非线性函数,因此对于平稳信号输入,采用BP神经网络构成自适应滤波器可以提高系统的抵消性能.根据神经元网络的自适应噪声抵消系统原理,通过仿真实验研究了在不同输入信噪比、不同通道函数、不同输入信号条件下系统的噪声抵消性能.实验表明BP方法噪声抵消效果显著,信噪比增益高.     

利用自适应滤波器的组合设计抵消多噪声

针对一个自适应滤波器不能解决多噪声信号的问题,提出了自适应滤波器的组合设计思想。在线性组合条件及LMS算法下,分析多噪声抵消原理,并进行多噪声语音信号的Matlab仿真实验,可以得到清晰的语音信号。结果表明,自适应滤波器的组合设计可以有效抵消多种干扰信号。     

基于自适应模糊神经网络的多传感器噪声抵消器

针对同一噪声源的多传感信号，采用自适应模糊神经网络系统(AFNNS)设计自适应噪声抵消器．采用AFNNS获取多路信息融合的权系数和自适应噪声抵消器的系数，基于AFNNS的自适应噪声抵消器不仅能获取信号的最佳估计，并且能克服模型和噪声存在的不确定性和不完备性．仿真结果表明，该自适应噪声抵消器的设计方法简单易行，去噪声效果优于基于平均法的去噪效果．     

基于形态学的ECG小波自适应去噪算法

为了消除心电信号中的噪声,提高心电监护仪的性能和计算机自动诊断效率,已经提出了多种方法来消除这些噪声.针对常规的ECG信号去噪算法存在的缺陷,提出了一种基于形态学的小波自适应去噪算法.该算法利用线性组合形态学滤波器去除基线漂移信号,然后对处理后无基漂的信号送入小波自适应滤波器,选取合适的阚值对其进行二次滤波去噪,最后得到无噪声的ECG信号.实验结果表明,该算法是一种有效的去噪算法.     

过程检测控制系统中的自适应噪声抵消法

文中分析了经典LMS噪声抵消法的局限性,讨论了过程检测系统信号和噪声的一般模型,并针对强噪声情况,提出了一种可行的噪声抵消法——基于建立相关模型的方法,该方法是利用自适应横向滤波算法。     

基于ADALINE神经网络的自适应滤波方法

自适应滤波器能够适应系统和环境的动态变化,具有较高的滤波精度。介绍了一种利用ADALINE神经网络进行自适应滤波的方法,根据自适应噪声抵消原理建立了ADALINE自适应神经滤波器模型,并使用该模型将发动机高压油管振动信号中的机体振动噪声滤除,提高了信噪比,为利用高压油管振动信号进行喷油器故障的精确诊断奠定了基础。     

基于形态学与小波变换的信号自适应去噪算法

针对常规的ECG（electrocardiogram）信号去噪算法存在的缺陷,提出了一种基于形态学与小波变换的自适应综合去噪算法。该算法利用形态学滤波器去除基线漂移信号,用小波滤波器去除高频干扰信号,并将这两部分所得到的心电噪声分量作为自适应滤波器的参考输入信号,对ECG信号进行自适应滤波处理,最后得到去噪后的ECG信号。实验表明,本算法是一种有效的去噪算法。     

基于卷积自编码神经网络的心电信号降噪

心电信号由于在采集过程中会受到外界环境的干扰导致其形态特征被严重淹没,从而对医生的诊断和远程智能分析造成干扰。基于此,提出了一种基于卷积自编码神经网络的心电信号降噪算法。该方法利用自编码器的编码、解码特性,通过卷积的方法构建深层神经网络来学习从含噪心电信号到干净心电信号的端对端映射。卷积层捕获心电信号的细节特征,同时消除噪声;解码部分能够对特征图进行上采样并恢复心电信号细节,从而得到干净的心电信号。实验中采用信噪比和均方根误差为指标,将该方法与小波阈值法、S变换法、BP神经网络法和指导滤波法进行比较。实验结果表明,该降噪方法整体降噪精度更优,同时信号的低频成分也得到了很好的保持。该方法可做到在消除心电信号中复杂噪声的同时完整保留心电信号的形态,为心血管疾病的智能诊断和心电图的特征检测奠定了基础。     

An improved method for on-line averaging and detecting of ECG waveforms

The most widely used methods for accurate signal averaging were studied and compared in order to gain a better understanding of the qualities and performances of each method. The level-triggering, contour-limiting, and correlation methods were simulated and compared. A new correlation method which is a weighted correlation of differences proved to be most suitable for real-time signal averaging, and detection of waveforms' variations. Simulated ECG waveforms and real ECG recordings were analyzed in this study. Twenty-eight ECG recordings of unipolar leads for noninvasive detection of the His-Purkinje activity were averaged separately by each method. The success in detection and the signal to noise ratio of the detected His activity obtained by each method was compared. Simulated ECG waveforms with random noise added were analyzed by four methods and the correct alignment as a function of the noise level was measured. The performance of our method in rejection of noisy waveforms and in detection of small variations in the waveforms is demonstrated.     

Improvements on self-adaptive voice activity detector for telephone data

Voice activity detection (VAD) has been studied for many decades and energy VAD is most commonly used. Energy VAD performs well under noise-free environments but deteriorates under noisy environment. Self-adaptive VAD performs much better than the traditional energy VAD in many aspects. However, one issue is that, the single one minimum energy threshold of the self-adaptive AVD could not perform well under the conditions with different channel varieties or background noises. In this paper, we make several improvements on the self-adaptive VAD to deal with that issue and enhance the detection performances. A k-means based average energy clustering approach is proposed to find better minimum energy thresholds for each speech recording. In the VAD decision phase, the new threshold is used for the likelihood ratio test. Furthermore, better results have been achieved by applying the median filtering as a post-processing step of self-adaptive VAD to smooth the short-time noise VAD errors. Experimental results on a subset of the NIST 2006 speaker recognition evaluation dataset show that our proposed method outperforms both the traditional energy-based and self-adaptive VAD approaches.     

Motion artifact removal based on periodical property for ECG monitoring with wearable systems

Motion artifact removal (MR) is one of the essential issues in processing raw ECG signals since it could not be simply solved by using classic filtering. In this paper, a QRS detection based Motion Artifact Removal algorithm (QRSMR) is proposed. The proposed method detects the entire QRS complex and removes the noise between two QRS complexes, while recovering P and T-waves. As verified in the tests on simulated noisy ECG signals, QRSMR outputs with seriously contaminated ECG signals have an increase of the correlation with their original clean signals from 40% to nearly 80%, demonstrating the improved noise removal ability of QRSMR. Moreover, in the tests on real ECG signals measured on volunteers with a flexible wearable ECG monitoring device developed at Fudan University, QRSMR is able to recover P-wave and T-wave from the contaminated signal, which shows its enhanced performance on motion artifact reduction comparing with adaptive filtering method and other methods based only on empirical mode decomposition.     

Frequency-domain digital filtering techniques for the removal of powerline noise with application to the electrocardiogram

This paper presents two new local processing frequency-domain methods for the removal of powerline noise from electrophysiological signals. The first is based on an iterative division or a multiplication of a set of frequencies centered at 60 Hz. The second users a basic property of the natural logarithm to smooth the 60-Hz noise. Both methods are intended to reduce powerline noise without affecting the frequency spectrum of the signal in the regions surrounding 60 Hz. For illustration, these local processing methods are applied to artificial and real electrocardiographic (ECG) data and are compared to a fixed IIR notch digital filter which is designed by pole-zero placements on the unit circle. The performance of each method is measured by the error squared, which is the square of the difference between the original noise-free signal and the filtered noisy ECG. Finally, since the two methods are iterative processes, comparison of their rate of convergence to a predefined noise reduction level is considered.     

低信噪比下噪声抑制的语音活动检测

传统的语音活动检测的方法,在噪声比较恶劣(一般指信噪比在5db以下)的环境下,效果很差,而本文提出的语音活动检测的方法在低信噪比的情况下仍然能够达到很好的效果.该方法主要包含两部分:第一部分是噪声抑制,第二部分是基于状态机的语音活动检测.通过实验结果可以证明,本文提出的方法在白噪声,嘈杂人声和汽车噪声的环境下比G.729采用的语音活动检测的方法提高很多.     

Robust vocabulary recognition clustering model using an average estimator least mean square filter in noisy environments

Noise estimation and detection algorithms must adapt to a changing environment quickly, so they use a least mean square (LMS) filter. However, there is a downside. An LMS filter is very low, and it consequently lowers speech recognition rates. In order to overcome such a weak point, we propose a method to establish a robust speech recognition clustering model for noisy environments. Since this proposed method allows the cancelation of noise with an average estimator least mean square (AELMS) filter in a noisy environment, a robust speech recognition clustering model can be established. With the AELMS filter, which can preserve source features of speech and decrease the degradation of speech information, noise in a contaminated speech signal gets canceled, and a Gaussian state model is clustered as a method to make noise more robust. By composing a Gaussian clustering model, which is a robust speech recognition clustering model, in a noisy environment, recognition performance was evaluated. The study shows that the signal-to-noise ratio of speech, which was improved by canceling environment noise that kept changing, was enhanced by 2.8 dB on average, and recognition rate improved by 4.1 %.     

一种新的自适应网络入侵检测模型

在基于免疫模型的网络入侵检测中,因模型对自体的动态变化缺乏自适应性导致高的误报率和漏报率。为了提高网络入侵检测模型在动态环境下的自适应性,使模型能更好地应对不断变化的外部环境,提出了一种新的自适应网络入侵检测模型。模型中详细阐述了自体的演化,对现有否定选择模型中检测器生成存在问题进行了分析,提出新的检测器生成算法,随着自体的在线自动更正,检测器可以始终保持同步更新。结果表明该模型具有很好自适应性和动态性,可以对入侵行为进行有效的识别。     

基于GF（28）域相似函数的椒盐噪声滤波算法

针对数字图像采集及处理过程中广泛存在的椒盐噪声,基于GF（2⁸）有限域,提出一种具有噪声点精确检测能力的改进型噪声滤波算法。根据图像相邻像素相关性,构建了GF（2⁸）域上的相似函数,并据此提出了GF（2⁸）域上的图像自适应椒盐噪声检测算子,可对噪声点准确定位;结合中值滤波算法完成数字图像的椒盐噪声自适应滤除。实验表明提出算法对噪声点定位准确,在图像的恢复和保护图像细节方面对比其他算法有较大提高,对强噪声污染图像的恢复也有较好效果。     

车载环境下的语音端点检测

为了提高车载噪声环境下语音端点检测的准确性,提出了一个基于GRU RNN的神经网络结构, 对带噪语音的Log Mel特征序列进行处理,实现语音与噪声的分离,从而恢复出纯净语音的Log Mel特征序列;在此基础上,提出一种新的特征Log Mel Sum,并用该特征进行端点检测。实验结果表明,在车载环境下,本文方法具有很好的端点检测性能。