本文選題：壓縮感知　切入點(diǎn)：MIMO雷達(dá)　出處：《南京信息工程大學(xué)》2017年碩士論文

【摘要】：MIMO雷達(dá)作為一種新體制雷達(dá),與傳統(tǒng)的相控陣?yán)走_(dá)相比,在目標(biāo)檢測、波束形成和參數(shù)估計等方面具有明顯的優(yōu)勢。在實(shí)際MIMO雷達(dá)應(yīng)用中,目標(biāo)往往只占據(jù)少數(shù)的分辨單元,即MIMO雷達(dá)的目標(biāo)回波信號是稀疏的。因此,壓縮感知理論能應(yīng)用于MIMO雷達(dá)的目標(biāo)參數(shù)估計問題中。本文設(shè)計了一種抗噪聲能力強(qiáng)的自適應(yīng)正則化SL0算法,以及研究MIMO雷達(dá)在病態(tài)感知矩陣和陣元失效條件下的多維目標(biāo)參數(shù)估計問題,主要內(nèi)容如下:(1)針對快速稀疏重構(gòu)算法—SL0算法的抗噪聲能力和穩(wěn)健性較差的問題,提出一種自適應(yīng)正則化的SL0算法。該算法在SL0算法的內(nèi)循環(huán)最速上升法中以第一次迭代的信號殘差項(xiàng)估計值以及該迭代前后的稀疏信號估計的偏差值作為當(dāng)前正則化參數(shù)的選擇依據(jù),從而能自適應(yīng)地調(diào)整在外循環(huán)迭代中的信號稀疏度和誤差容許項(xiàng)的權(quán)重值,在優(yōu)化過程中保持兩者的平衡性,從而有效降低稀疏信號的重構(gòu)誤差,提高了 SL0算法的抗噪聲干擾能力。(2)針對MIMO雷達(dá)因感知矩陣病態(tài)而導(dǎo)致SL0算法失效的問題,利用修正截斷奇異值分解方法改善MIMO雷達(dá)的病態(tài)感知矩陣,使得SL0算法能有效應(yīng)用于MIMO雷達(dá)的快速多目標(biāo)參數(shù)估計。為了方便科研人員測試MIMO雷達(dá)的目標(biāo)參數(shù)估計性能,開發(fā)了基于LabVIEW的病態(tài)感知矩陣下MIMO雷達(dá)目標(biāo)參數(shù)估計測試軟件。(3)為了解決MIMO雷達(dá)因陣元失效而導(dǎo)致其目標(biāo)參數(shù)估計性能下降的問題,將矩陣填充應(yīng)用于陣元失效MIMO雷達(dá)的目標(biāo)參數(shù)估計。在失效陣元輸出的整行或整列零元素上疊加微小的服從高斯分布的隨機(jī)擾動量,使其能滿足矩陣填充條件,并利用矩陣填充和迭代加權(quán)l(xiāng)q方法獲得目標(biāo)場景向量粗估計值,然后根據(jù)目標(biāo)場景向量粗估計值和感知矩陣重構(gòu)出失效陣元的目標(biāo)接收數(shù)據(jù),從而相比于未利用矩陣填充的重構(gòu)算法,能以較高精度估計出目標(biāo)的三維參數(shù)。
[Abstract]:As a new type of radar, MIMO radar has obvious advantages in target detection, beamforming and parameter estimation, compared with traditional phased array radar. In practical applications of MIMO radar, the target usually occupies only a few resolution units. That is, the target echo signal of MIMO radar is sparse. Therefore, compression sensing theory can be applied to the estimation of target parameters in MIMO radar. In this paper, an adaptive regularization SL0 algorithm with strong anti-noise ability is designed. And the multi-dimension target parameter estimation problem of MIMO radar under the condition of ill-conditioned perception matrix and array element failure is studied. The main contents are as follows: (1) aiming at the problem of poor anti-noise ability and robustness of the fast sparse reconstruction algorithm -SL0 algorithm, An adaptive regularization SL0 algorithm is proposed, in which the estimation of the signal residual of the first iteration and the deviation of the sparse signal before and after the first iteration are taken as the current positive values in the SL0 algorithm. The basis for the selection of chemical parameters, Thus, the signal sparsity and the weight of the error tolerance can be adjusted adaptively in the outer loop iteration, and the balance between the two can be maintained in the optimization process, thus effectively reducing the reconstruction error of the sparse signal. In order to solve the problem of SL0 algorithm failure caused by ill-conditioned perception matrix of MIMO radar, the modified truncated singular value decomposition (SVD) method is used to improve the ill-conditioned perceptual matrix of MIMO radar. The SL0 algorithm can be applied to the fast multi-target parameter estimation of MIMO radar effectively. In order to facilitate the researchers to test the performance of MIMO radar target parameter estimation, In order to solve the problem that the target parameter estimation performance of MIMO radar is degraded due to the failure of array elements, a testing software of MIMO radar target parameter estimation based on LabVIEW ill-conditioned perception matrix is developed in order to solve the problem that the performance of MIMO radar target parameter estimation is degraded due to the failure of array elements. The matrix filling is applied to estimate the target parameters of the array element failure MIMO radar. The random perturbation quantity distributed from Gao Si is superimposed on the whole line or whole column zero element of the invalid array element, so that the matrix filling condition can be satisfied. The rough estimate of target scene vector is obtained by filling matrix and iterative weighted LQ method, then the target receiving data of invalid matrix element is reconstructed according to the coarse estimation value of target scene vector and perception matrix. Compared with the reconstruction algorithm without matrix filling, the 3D parameters of the target can be estimated with high accuracy.
【學(xué)位授予單位】：南京信息工程大學(xué)
【學(xué)位級別】：碩士
【學(xué)位授予年份】：2017
【分類號】：TN958

【參考文獻(xiàn)】

相關(guān)期刊論文 前5條

1 馮俊杰;張弓;文方青;;基于SL0范數(shù)的改進(jìn)稀疏信號重構(gòu)算法[J];數(shù)據(jù)采集與處理;2016年01期

2 王超宇;賀亞鵬;胡恒;朱曉華;;基于貝葉斯壓縮感知的噪聲MIMO雷達(dá)目標(biāo)成像[J];南京理工大學(xué)學(xué)報;2013年02期

3 王軍華;黃知濤;周一宇;;稀疏信號重構(gòu)的迭代平滑l_0范數(shù)最小化算法[J];宇航學(xué)報;2012年05期

4 顧福飛;池龍;張群;彭發(fā)祥;朱豐;;基于壓縮感知的稀疏陣列MIMO雷達(dá)成像方法[J];電子與信息學(xué)報;2011年10期

5 楊明磊;陳伯孝;秦國棟;張守宏;;多載頻MIMO雷達(dá)的空時超分辨算法[J];電子與信息學(xué)報;2009年09期

，

本文編號：1672235