基于SVD与数学形态学分形维数谱的战场声特征提取

分形维数作为战场声信号的特征,存在特征数量不足,反映信号非线性不充分的问题,提出了一种基于SVD与数学形态学分形维数谱(Singular Value Decomposition And Mathematical Morphological Fractal Dimensions Spectrum,SVD-MMFDS)的战场声特征提取方法。对声信号构造Hankel矩阵,再进行SVD分解,根据信号频率与奇异值的关系,重构信号分量。将这些重构信号依次线性叠加,每叠加一次信号分量就计算一次分形维数,直至完全恢复原信号;通过这种方法,构成数量多且更能反映信号非线性的分形维数谱。运用半实物仿真实验将SVD与数学形态学分形维数谱的方法,与变分模态分解(VMD)和分形维数结合的方法进行对比,该方法提取的战场声特征具有更好的区分度且特征数量更多,为利用信号非线性来识别战场声目标提供较好的选择。

飞机声信号的特征提取与识别

在对某型喷气飞机声信号做时频域分析的基础上,采用LMS自适应横向滤波对飞机声信号作预处理,而后分别采用过零率和小波变换作目标特征提取。对这两种方法提取的特征再分别应用BP神经网络作目标分类识别,结果表明这两种特征提取方法与BP神经网络分类器相结合对处于起飞或降落滑跑中的喷气飞机均有较高的识别率。

声发射信号特征提取方法在复合材料层合板中的应用

将小波包能量谱分析法应用到碳纤维复合材料层合板(CFRP层合板)中,对CFRP层合板的落锤冲击实验过程进行分析,得到CFRP层合板在冲击实验时的损伤类型和各损伤阶段所呈现出来的声发射信号特征,实现材料损伤类型的判别。

多重分形的耐火材料损伤声发射信号特征提取方法

为了对镁碳质耐火材料损伤声发射信号进行有效的特征提取,研究了利用非线性领域的分形方法进行特征提取的可行性。首先利用仿真声发射信号研究了多重分形谱参数(Δα、Δf、K),广义分形维数均值Mean Dq等参数表征信号频谱结构特征的能力,结果表明多重分形谱宽Δα能够有效的表征信号的频谱特征;然后对镁碳质耐火材料进行单轴压缩试验,采集损伤声发射信号,并计算出两种典型损伤声发射信号的多重分形谱宽Δα,通过多重分形谱宽的分布证实了利用其进行特征提取是可行的。

无人机飞行声特征与提取方法比较

目前,对无人机飞行声信号的分析主要是基于传统语音信号处理的手段,并未进行深入分析。文中针对无人机飞行声信号,结合无人机的气动特点深入研究分析得出无人机声信号的特征,分析比较傅里叶变换(FFT)、梅尔倒谱系数(MFCC)和基因周期3种特征提取算法并提取特征,应用支持向量机(SVM)分类算法,进行机型识别分类。实测与实验结果表明,FFT与MCC识别率相近,FFT运算复杂度低,基因周期不太适合单独进行特征识别,因此得出FFT适合作为无人机声特征提取方法。

大型网箱养殖鱼群声信号短时频谱特性

【目的】研究大型网箱养殖鱼群声信号特性,为鱼类声学特性研究和智能养殖提供参考。【方法】利用被动声学技术对大型网箱鱼群进行昼夜持续监测,计算人工投料、自动投料、摄食后以及深夜安静等特殊时段下鱼群发声信号的短时谱,获取短时能量和短时过零率的统计特征,对比区分网箱混叠三种鱼群在不同外界激励条件下的发声频段、发声方式和短时特征。【结果与结论】1)石斑鱼(Epinephelus drummondhayi)进食时发出咕噜和轰隆声,信号频带为200~1000 Hz;卵形鲳鲹(Trachinotus ovatus)进食过程通过喉齿摩擦和拍打发出声音,信号频带为600~2000 Hz;叫姑鱼(Johnius grypotus)摄食时通过鳔内的气体振动发出以200~1500 Hz为基频的声音,且存在持续到7000 Hz的高频谐波。2)鱼群游动身体相互摩擦和拍打水花的噪声频段为2000~5000 Hz。3)网箱混叠三种鱼群的声信号在短时特征上可得到区分,短时特征值大小为叫姑鱼>卵形鲳鲹>石斑鱼。

Classification of underwater still objects based on multi-field features and SVM

A Support Vector Machine is used as a classifier to the automatic detection and recognition of underwater still objects. Discrimination between the objects can be transferred into different projection spaces by the process of multi-field feature extraction. The multi-field feature vector includes time-domain, spectral, time-frequency distribution and bi-spectral features. Underwater target recognition can be considered as a problem of small sample recognition. SVM algorithm is appropriate to this kind of problems because of its outstanding generalizability. The SVM is contrasted with a Gaussian classifier and a k-nearest classifier in some experiments using real data of lake or sea trial. The experimental results indicate that SVM is better than the others two.

Classifying ships by their acoustic signals with a cross-bispectrum algorithm and a radial basis function neural network

An algorithm for estimating the cross-bispectrum of an acoustic vector signal was formulated. Composed features of sound pressure and acoustic vector signals are extracted by the proposed algorithm and other estimating algorithms for secondary and higher order spectra. Its effectiveness was tested with lake and sea trial data. These features can be used to construct an input vector set for a radial basis function neural network. The classification of vessels can then be made based on the extracted features. It was shown that the composed features of acoustic vector signals are more easily divided into categories than those of pressure signals. When using the composed features of acoustic vector signals, the recognition rate of underwater acoustic targets improves.

Research of multi-channel impulse noise detection algorithms based on template

Attenuating the undesired audio noise generated by impulse noise,such as shot and scream of brakes,is specially useful for real-time audio recording of TV or broadcasting live report.On the basis of impulse noise detection algorithms based on template,this paper improves the method of establishing the template by using multiple microphones to pick up noise corrupted signals and impulse noises in the environment.The universal of thresholds is found and a detection algorithm with slope as the characteristic is proposed by comparing a variety of feature extraction algorithms.The proposed algorithm gets a significant improvement in testing speed and accuracy,which means it is suitable for real-time processing of audio signals.

Intrinsic mode characteristic analysis and extraction in underwater cylindrical shell acoustic radiation

Target dimension is important information in underwater target classification. An intrinsic mode characteristic extraction method in underwater cylindrical shell acoustic radiation was studied in this paper based on the mechanism of shell vibration to gain the information about its dimension instead of accurate inversion processing. The underwater cylindrical shell vibration and acoustic radiation were first analyzed using mode decomposition to solve the wave equation. The characteristic of acoustic radiation was studied with different cylindrical shell lengths, radii, thickness, excitation points and fine structures. Simulation results show that the intrinsic mode in acoustic radiation spectrum correlates closely with the geometry dimensions of cylindrical shells. Through multifaceted analysis, the strongest intrinsic mode characteristic extracted from underwater shell acoustic radiated signal was most likely relevant to the radiated source radius. Then, partial information about unknown source dimension could be gained from intrinsic mode characteristic in passive sonar applications for underwater target classification. Experimental data processing results verified the effectiveness of the method in this paper.