基于高阶谱的舰船辐射噪声特征提取

在对现有舰船辐射噪声信号分布研究的基础上,使用可抑制高斯噪声的高阶谱（双谱和三谱）对其特征进行了分析。分析结果表明,不同类的舰船,其主频范围有其固定的区域。Ｉ类船其主频一般在１２０Ｈｚ以下,Ⅱ类船一般在１３０～３２０Ｈｚ之间。对于双谱图,不同的ω１与ω２的关系,其频谱特征有所不同,对于三谱也有同样的结果。本文中双僧取ω１＝ω２,三谱取ω１＝ω２＝ω３两种关系,以它们最大谱值对应的ω１以及Ｂ（ω１,ω２）中极大值所对应的ω１、ω２组成特征向量．     

舰船辐射噪声的高阶统计量特征提取及特征压缩

提取了6类水中目标的11/2谱统计线谱特征和AR模型特征。实验结果表明基于11/2谱的统计线谱特征和AR模型特征有效的识别出了目标,达到了85%的综合识别率。最后利用主元分析技术对上述特征进行了压缩,在保持识别率的同时将特征维数从64维压缩至34维,有利于工程上的实现。     

非线性时间序列的高阶奇异谱分析

基于反映线性相关结构的协方差矩阵的奇异谱分析，本质上是一种线性的方法.奇异谱分析用于吸引子重构的可靠性问题引发了一些争议.本文基于具有盲高斯噪声及体现非线性相关等性质的高阶累积量，提出了一种高阶的奇异谱分析方法.通过对Hénon映射、Logistic映射和Lorenz模型的分析说明了该方法的有效性，并在不同的延时、嵌入维数、抽样时间及有噪声的情况下表现出较好的鲁棒性. 关键词：     

船舶辐射噪声模式识别

本文用模式识别技术,对航道上航行的船舶进行分类判决.根据船舶辐射噪声的二阶统计特性——自相关函数,我们提取了四个特征量,用统计判庆理论对三类船舶的辐射噪声进行了分类判决.文中给出了海上实验数据的分类结果.     

高阶相对论谐波辐射的理论研究

对短脉冲强激光在非稠密均匀等离子体中传播时产生的相对论相干谐波辐射进行了一般性研究．在准稳态近似（ｑｕａｓｉｓｔａｔｉｃ ａｐｐｒｏｘｉｍａｔｉｏｎ）下得出ｎ阶谐波包络的演化方程，并用此方程研究了ｎ阶谐波的增长与饱和及转化效率等问题     

舰船辐射噪声的动态识别

以舰船辐射噪声的包络谱分析为基础,讨论了舰船包络谱的特点和提取的方法。据此提出了一种新的目标识别方法。该方法能有效提高舰船噪声识别系统的识别准确率,还可同时获得有关目标舰船的重要运动要素。     

舰船辐射噪声的子波分析

本文把子波变换多分辨的特点应用于舰船噪声的分析中。这种新的分析方法克服了短时间谱在时频平面上均匀分割带来低频域频率分辨率低,而高频域时间分辨低的缺点。船辐射噪声的尺度谱保留了传统倍频程方法在频率域分割的优点,且还有高频域时间分辨高的特点。从大庆31号油船辐射噪声的尺度谱上可清楚看到在低频域上有二根横线,其尺度值换算为频率后为27.8 Hz和44.1 Hz,反映在此频率附近,信号能量较高;同时还可清楚看到在高频域上有9.5 Hz叶片频的调幅现象。在它的短时谱图中是无法同时看到这二个特征的。本文在描述和显示船辐射噪声的谱特征时充分利用它们符合局部平稳过程的特点,把平稳谱和非平稳谱拆开描述和显示,获得很好的结果     

船舶辐射噪声的混沌现象研究

水下弱信号的检测和识别是当今水声信号处理领域中存在的难题.应用混沌、分形理论,从根空间轨迹、Lyapunov指数和关联维等方面研究了船舶辐射噪声的混饨现象,发现船舶辐射噪声信号的确存在混沌吸引子,且不同类别的信号具有不同的吸引子维数.这一结果将为水声信号处理,为水下目标探测和识别提供崭新的理论手段。     

舰船辐射噪声的分形布朗运动模型

通过对分形布朗运动的研究,建立了舰船辐射噪声的分形布朗运动模型。在此模型基础上,提取了舰船辐射噪声的分数维参数,并构造了基于分形布朗随机场的分形特征矢量,在水声目标识别系统中提供了一种新的特征分析手段、通过分类对比实验,表明分形特征矢量可以成为水声目标分类的重要依据,能够有效地补充其他特征提取方法的不足.     

船舶辐射噪声的自然尺度特征

船舶辐射噪声是非常复杂的,寻找新的特征是目前水下目标识别中的一项非常迫切而艰巨的任务。本文在研究了相空间重构技术的基础上,基于混沌信号的自然尺度提出了一种新的船舶辐射噪声的特征提取算法,此特征在分类识别实验中获得了比较好的效果。实践证明,船舶辐射噪声的自然尺度特征是一种比较有效的特征,它不仅具有较好的实时性,而且具有较高的识别率。     

基于深度学习的船舶辐射噪声识别研究

为了改善船舶辐射噪声识别系统的性能,进一步提高船舶辐射噪声识别的正确率,该文提出采用一种基于深度学习的船舶辐射噪声识别方法。该方法首先提取了船舶辐射噪声的频谱、梅尔倒谱系数等特征,将提取特征后的图像样本分别用于训练卷积神经网络和深度置信网络,再对船舶辐射噪声进行识别。通过文中所给实例,将深度学习和支持向量机两种识别方法的性能进行比较,得出深度学习方法可以有效地提高船舶辐射噪声识别正确率的初步结论。     

Motor bike piston-bore fault identification from engine noise signature analysis

Imitation of the act of diagnosing engine faults by an expert auto-mechanic just by hearing the noise from a defective vehicle has been attempted with a robust instrumentation technique. In the present experimental work, the prime objective is to establish a process to identify the piston-bore defect by analyzing the engine noise. The aim is to develop a robust filtering algorithm in order to be able to use the technique in the natural environment of an auto workshop. The algorithm uses engine noise data from healthy and defective vehicles acquired in the natural workshop environment. Effectiveness of conventional parameters that are used (six of the most preferred) in statistical learning systems to tackle similar problems has been assessed. Additionally, six more statistical parameters have been derived and proposed to achieve a more effective statistical-learning based decision making system. These additional parameters have been derived from the continuous wavelet transform (CWT) coefficients. The appropriate coefficient level has been decided by using the frequency marginal integration of the CWT coefficients. A deeper level of CWT scalogram analysis has been carried out to establish the appropriateness of complex Morlet wavelet function, for the class of problem under investigation. Then the proposed technique uses a third degree polynomial kernel function and sequential optimization based support vector machine (SVM) based on all the twelve parameters in order to identify the piston-bore defect from engine noise signature. It has also been shown that Feed-forward Back-propagation Neural Network (FBNN) is equally effective with higher number of training samples.     

一种提取光反馈混沌激光器时延信息的谱分析方法

鉴于时延信息严重威胁着混沌通信的保密性,提出了一种利用频谱仪直接确定激光器时延信息的新方法。通过记录和观察信号的频谱,发现其频谱受到了时延信息的调制;进一步对频谱作逆傅里叶变换,即可直接获取光反馈混沌激光器的时延信息。实验中对腔长L=11.02 m的单反馈和腔长近似相等(L1=11.02 m, L2=11.25 m)或腔长成整数比(L1=11.02 m, L2=22.03 m) 的双反馈情形进行了研究。分析结果表明:在两外腔长度近似相等的情况下,2个外腔长度信息不能被隐藏;当反馈外腔长度成整数比时可以从频谱中提取出半导体激光器的反馈延迟时间。     

Describing supercontinuum noise and rogue wave statistics using higher-order moments

We show that the noise properties of fiber supercontinuum generation and the appearance of long-tailed “rogue wave” statistics can be accurately quantified using statistical higher-order central moments. Statistical measures of skew and kurtosis, as well as the coefficient of variation provide improved insight into the nature of spectral fluctuations across the supercontinuum and allow regions of long-tailed statistics to be clearly identified. These moments – that depend only on analyzing intensity fluctuations – provide a complementary tool to phase-dependent coherence measures to interpret supercontinuum noise.     

Traffic noise spectrum analysis: Dynamic modeling vs. experimental observations

This paper compares two traffic representations for the assessment of urban noise frequency spectrum: (i) a static one, based on mean vehicle speeds and flow rates, (ii) a dynamic one, which considers vehicle interactions along the network. The two representations are compared on their suitability to match real on-field noise levels, recorded on a three lane quite busy street. Representation (i) fails in reproducing spectra envelopes that correspond to this site. In particular, it underestimates low frequencies, what can conceal the real impact of traffic flow on urban sound quality. Representation (ii) greatly improves estimation. It guarantees accurate environmental noise assessment, since it reproduces all traffic situations that are encountered in the site. Moreover, its 1s-based structure allows for the evaluation of spectra variations, with a good accuracy.     

Theoretical analysis of the ship noise demodulation spectrum harmonic clan feature

The structure law of the ship noise demodulation spectrum harmonic clan feature and corresponding applications to the recognition problem are studied by theoretical derivation and sample analysis.In order to recognize the propeller blade-number,the ship propeller cavitation noise model is established on the basis of five hypotheses.Further,the mathematical expressions of the structure law are deduced by the model.And the class space of the propeller blade-number is analysed mathematically.The results can be applied to target recognition as expert knowledge.Finally,the demodulation spectrum harmonic clans of experiment samples are obtained by using the wavelet packet.The structure law is validated by the statistic analysis of experiment samples.It is shown that the statistical average results of experiment samples are consistent very well with the theory structure law,and 78.6% of the samples basically accord with the theory structure law.     

The spectrum of combustion-generated noise

A calculation of the energy release rate resulting from the combustion of propane-air mixtures is presented and the result is used to calculate the far field noise spectrum for an open flame by using appropriate Fourier transform techniques. The results illustrate the broad band nature of combustion noise and show that, for the range of parameters indicated, the peak frequency in the $\text{1}\text{3}$ octave band is in the range 400–1000 Hz. The results also indicate that the shape of spectrum is influenced by the time history of the heat release rate and the turbulence intensity and length scales; on the other hand, the peak frequency is a function of the heat release per unit mass of fuel which is essentially the same for hydrocarbon fuels.     

基于背景噪音谱分析的匹配滤波器设计研究

在空间匹配滤波器设计过程中,为抑制输入场景中背景噪音对输出相关平面上相关点的影响,通过比较匹配滤波函数和背景噪音的傅里叶谱,根据两者差值的量级对匹配滤波函数的振幅进行非线性调制,在匹配滤波过程中使设计的空间匹配滤波器对强噪音的傅里叶谱处透过率降低,而尽量保持输入图像傅里叶谱的高透过率,从而增加输出相关平面上的信噪比,实现相关点的正确探测。计算机模拟和实验结果表明：与纯相位匹配滤波器相比其抗噪能力得到较大改善。     

On-line noise source identification based on the power spectrum estimation and grey relational analysis

The identification of acoustic source accurately is a fundamental problem in noise control. In the practical project, if the contribution of multi-source-noise to the whole was identified, and then the noise level can be reduced accordingly. This paper presents a new approach to acoustic noise identification by introducing modern spectrum estimation and grey relational analysis (GRA). Modern spectrum was used to recognize the main noise source and GRA was used to recognize the similarity among different curves of power spectrum. The ranking of the noise sources was obtained on the basis of their individual contribution to the overall noise. The results of simulation signals confirmed the feasibility and validity of the method proposed in this dissertation and it will play an important role in noise control, signal source identification and other fields.