發(fā)布時間：2018-04-16 00:41

  本文選題：寬帶信號 + 降噪　； 參考：《電子科技大學》2014年碩士論文

【摘要】：寬帶信號無論在生活中還是軍事上的應用都越來越廣,但是在應用過程中,不可避免的會在信號中混入噪聲,嚴重的情況下甚至會讓信號面目全非。所以寬帶信號降噪就成為了一個很有研究價值和意義的課題。本文在對目前主流降噪算法進行了全面了解的基礎上,選擇了常應用于窄帶信號進行降噪的小波閾值降噪算法,對寬帶線性調頻(LFM)信號進行降噪的嘗試。并對這種算法從兩個方面進行了改進。進行了仿真對比,分析出每一種方法的優(yōu)缺點。最后將兩者結合成一種降噪方法,對寬帶LFM信號的降噪取得了較好的效果。本文的主要工作有:1.介紹小波分析的相關基本理論,由連續(xù)小波變換,二進小波變換,小波級數(shù)三個層次,遞進的說明小波分析對信號進行分解的原理,并詳細闡述了多分辨分析和mallat快速算法的過程。2.介紹了小波閾值降噪算法的原理和流程,對小波基的選取、閾值函數(shù)的選取和閾值估計,以及降噪以后效果的評估體系進行了詳細說明,并針對寬帶LFM信號選取了合適的參數(shù)。3.通過對常見的小波閾值降噪算法進行仿真對比,分析出這種方法的局限性,在低頻噪聲過濾和時間分割兩個方面進行了算法的改進,通過實驗仿真證明這種改進有較好的效果。
[Abstract]:Wideband signal is more and more widely used in life and military affairs, but in the process of application, it is inevitable to mix noise in the signal, and even make the signal completely different in serious cases.So wideband signal noise reduction has become a research value and significance of the subject.On the basis of a comprehensive understanding of the current mainstream denoising algorithms, the wavelet threshold de-noising algorithm, which is often applied to narrow band signals, is selected in this paper to try to reduce the noise of wideband linear frequency modulation (LFM) signals.The algorithm is improved from two aspects.The advantages and disadvantages of each method are analyzed.Finally, the two methods are combined into a noise reduction method, and good results are obtained for wideband LFM signal de-noising.The main work of this paper is: 1.This paper introduces the basic theory of wavelet analysis, which consists of three levels: continuous wavelet transform, dyadic wavelet transform and wavelet series.The process of multi-resolution analysis and mallat fast algorithm. 2. 2.The principle and flow of wavelet threshold denoising algorithm are introduced. The selection of wavelet basis, the selection of threshold function and threshold estimation, as well as the evaluation system of effect after denoising are described in detail, and the appropriate parameter. 3 is selected for wideband LFM signal.The limitation of this method is analyzed by comparing the common wavelet threshold denoising algorithms. The algorithm is improved in the aspects of low frequency noise filtering and time segmentation. The experimental results show that the improved algorithm has good results.
【學位授予單位】：電子科技大學
【學位級別】：碩士
【學位授予年份】：2014
【分類號】：TN911.4

【參考文獻】

相關期刊論文 前1條

1 吳俊明;;論小波分析理論的應用研究與發(fā)展前景[J];長春理工大學學報(綜合版);2005年02期

，

本文編號：1756564