基于最优匹配小波的舰船辐射噪声分析

发布时间：2018-06-24 00:45

本文选题：舰船辐射噪声 + 匹配小波　；参考：《昆明理工大学》2014年硕士论文

【摘要】：舰船辐射噪声是舰船特征信号的一个重要标志,其线谱由于传播距离远、衰减缓慢、能量集中、所含信息量丰富等特点成为研究舰船辐射噪声的重点。由于海洋背景噪声复杂多变,实际接收的舰船辐射噪声十分微弱,线谱变得越来越难于检测和提取,因此线谱识别问题值得深刻研究。小波分析是近年来发展起来的信号处理方法,其突出的时频局部特性及优异的去相关能力得到了广泛应用,并取得了显著的效果,特别是能有效检测信号的奇异点。本文在剖析舰船辐射噪声特点的基础上,从一个新的角度,根据舰船辐射噪声的频域特征,从选取最优小波基入手,由小波变换与奇异性的联系对低信噪比的舰船辐射噪声的线谱进行分析。按照小波基选取三个准则,选取bior5.5小波为最优基,并通过不同小波基对舰船辐射噪声功率谱奇异性的检测比较,验证了bior5.5的最优性,其近似正交性保证了重构信号的精确度,也兼顾了对小波基紧支性和正则性的要求。为了精确模拟低信噪比信号中的真实线谱,通过设置不同信噪比噪声背景下,将舰船辐射噪声分低频段、中频段、高频段三个频带进行多角度线谱分析。在低频段,本文论述了基于小波变换的信号匹配分析方法。经不同小波基处理的实验结果比较,只有bior5.5小波基能准确提取舰船辐射噪声的特征；在中频段,本文总结构造阈值函数用于小波去噪的必要条件下,改进了一种新阈值函数,并巧妙地将其应用到舰船辐射噪声分析,重新完善了基于小波变换的中频段信号匹配分析方法,实现了线谱提取；在高频段,则引入基于小波包变换的信号匹配分析方法,大量实验表明三种方法的有效性、可行性以及各自的匹配性。本文首次从频域的角度对舰船的特征信号进行小波分析,得到了一些有意义的初步结果,证明了bior5.5小波基对舰船辐射噪声信号分析有良好的分辨率。
[Abstract]:Ship radiated noise is an important symbol of ship characteristic signal. Because of its long propagation distance, slow attenuation, concentrated energy, abundant information and so on, its line spectrum has become the focus of research on ship radiated noise. Because of the complex and changeable background noise and the weak radiated noise of the ship actually received, the line spectrum becomes more and more difficult to detect and extract, so the problem of line spectrum recognition is worth studying deeply. Wavelet analysis is a signal processing method developed in recent years. Its outstanding time-frequency local characteristics and excellent ability of de-correlation have been widely used, and have achieved remarkable results, especially the singularity can be effectively detected. On the basis of analyzing the characteristics of ship radiated noise, from a new point of view, according to the frequency domain characteristics of ship radiated noise, this paper begins with the selection of optimal wavelet basis. The linear spectrum of ship radiated noise with low signal-to-noise ratio (SNR) is analyzed by the relation between wavelet transform and singularity. According to three criteria of wavelet basis selection, bior5.5 wavelet is chosen as the optimal basis, and the singularity of power spectrum of ship radiated noise is detected by different wavelet bases. The optimality of bior5.5 is verified, and its approximate orthogonality ensures the accuracy of reconstructed signal. It also gives consideration to the requirements of wavelet basis compactness and regularity. In order to accurately simulate the true line spectrum of low signal-to-noise ratio (SNR) signal, the ship radiated noise is divided into three frequency bands: low frequency band, middle frequency band and high frequency band by setting different signal-to-noise ratio (SNR) noise background. In low frequency band, the method of signal matching analysis based on wavelet transform is discussed in this paper. By comparing the experimental results of different wavelet bases, only bior5.5 wavelet bases can accurately extract the features of ship radiated noise. In the middle frequency band, this paper summarizes the necessary conditions of constructing threshold functions for wavelet denoising, and improves a new threshold function. It is skillfully applied to the analysis of ship radiated noise, and the matching analysis method based on wavelet transform is improved to realize line spectrum extraction, and the signal matching analysis method based on wavelet packet transform is introduced in high frequency band. A large number of experiments show that the three methods are effective, feasible and matching. In this paper, the characteristic signals of ships are analyzed by wavelet analysis in frequency domain for the first time, and some significant preliminary results are obtained. It is proved that bior5.5 wavelet bases have good resolution for the analysis of ship radiated noise signals.
【学位授予单位】：昆明理工大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：U674.70;TN911.4

【参考文献】