本文選題：高分辨率 + 散射中心��； 參考：《西安電子科技大學》2014年碩士論文

【摘要】：隨著理論研究的不斷深入和雷達信號分辨率的不斷提升,基于寬帶高分辨率雷達信號的目標識別成為了雷達信號處理領域的一個重要發(fā)展方向。寬帶高分辨率雷達獲取的目標回波不再是一個“點”,而是由一系列分布在不同距離單元上的目標散射點構成的高維信號。通過提取并分析目標回波中的散射點可以獲取反映目標本質屬性的特征,如目標體的物理尺寸以及典型部件的尺寸和結構等信息。本文正是基于寬帶高分辨率雷達回波數(shù)據(jù),研究了目標散射中心特征的提取方法。論文主要內(nèi)容概括如下:1.分析比較了四種稀疏求解算法,并引入交替投影法改進算法。首先介紹了四種稀疏求解算法的基本原理和操作流程,結合實驗比較了各個算法的應用效果。然后提出了一種結合交替投影法的稀疏求解方法,通過不同的實驗設置,驗證了所提方法能夠提高稀疏求解的精度,并在一定程度上改善了求解算法對噪聲的穩(wěn)健性。2.研究了一維簡單散射中心特征的提取方法。介紹了一維散射中心的幾何繞射模型,以及極化散射矩陣的基本概念;提出了一種聯(lián)合多極化通道高分辨率距離像的散射中心提取方法。通過仿真實驗證明,與傳統(tǒng)方法相比所提方法在目標散射中心位置和極化散射矩陣的估計方面有一定優(yōu)勢。3.研究了屬性散射中心提取方法。(1)提出了一種結合交替投影與正交匹配追蹤的屬性散射中心提取方法。首先提取目標的合成孔徑雷達(Synthetic Aperture Radar,SAR)圖像支撐區(qū)實現(xiàn)對目標參數(shù)的自動初始化,減少了字典維度,降低了計算復雜度;然后利用正交匹配追蹤法估計目標屬性散射中心參數(shù),最后通過交替投影法對所估參數(shù)進行修正。該方法能夠降低目標屬性散射中心間的高度耦合對散射中心參數(shù)估計的影響,提高了估計精度。(2)提出了一種結合圖像信息的頻域屬性散射中心特征提取方法。該方法在頻域方法的估計過程中,引入目標的圖像信息實現(xiàn)對散射中心間的去耦合,既降低了計算復雜度,也改善了參數(shù)估計精度。
[Abstract]:With the deepening of theoretical research and the improvement of radar signal resolution, target recognition based on wideband high resolution radar signal has become an important development direction in the field of radar signal processing. The target echo obtained by wideband high-resolution radar is no longer a "point", but a high-dimensional signal composed of a series of scattering points distributed on different distance units. By extracting and analyzing the scattering points in the echo of the target, the characteristics reflecting the essential properties of the target can be obtained, such as the physical size of the target body, the size and structure of the typical components, and so on. Based on the wideband high resolution radar echo data, this paper studies the extraction method of target scattering center feature. The main contents of this paper are summarized as follows: 1: 1. Four sparse solving algorithms are analyzed and compared, and the alternating projection method is introduced to improve the algorithm. Firstly, the basic principle and operation flow of four sparse solving algorithms are introduced, and the application results of each algorithm are compared with experiments. Then a sparse solution method combined with alternating projection method is proposed. Through different experimental settings, it is verified that the proposed method can improve the accuracy of sparse solution, and improve the robustness of the algorithm to noise to a certain extent. A method for extracting the feature of one-dimensional simple scattering center is studied. The geometric diffraction model of one-dimensional scattering center and the basic concept of polarimetric scattering matrix are introduced, and a method of extracting scattering center with high resolution range profile of multi-polarization channel is proposed. The simulation results show that the proposed method has some advantages over the traditional method in estimating the location of the target scattering center and the polarization scattering matrix. In this paper, the method of extracting attribute scattering center is studied. (1) an attribute scattering center extraction method combining alternating projection and orthogonal matching tracking is proposed. Firstly, the SAR synthetic Aperture radar (SAR) image support region of the target is extracted to realize the automatic initialization of the target parameters, which reduces the dictionary dimension and computational complexity, and then estimates the scattering center parameters of the target attributes by orthogonal matching tracking method. Finally, the estimated parameters are modified by alternating projection method. This method can reduce the influence of high coupling between scattering centers of target attributes on the estimation of scattering center parameters and improve the estimation accuracy. In the estimation process of frequency-domain method, the image information of the target is introduced to realize the decoupling between scattering centers, which not only reduces the computational complexity but also improves the precision of parameter estimation.
【學位授予單位】：西安電子科技大學
【學位級別】：碩士
【學位授予年份】：2014
【分類號】：TN957.51

【參考文獻】

相關期刊論文 前1條

1 杜小勇;胡衛(wèi)東;郁文賢;;基于稀疏成份分析的幾何繞射模型參數(shù)估計[J];電子與信息學報;2006年02期

相關博士學位論文 前1條

1 杜蘭;雷達高分辨距離像目標識別方法研究[D];西安電子科技大學;2007年

，

本文編號：2010093