【摘要】：認(rèn)知雷達(dá)(Cognitve Radar,CR)是一種新型的雷達(dá)體制,具有學(xué)習(xí)、理解和推斷環(huán)境的能力,具備接收機(jī)向發(fā)射機(jī)的持續(xù)信息反饋和波形自適應(yīng)發(fā)射能力。自適應(yīng)發(fā)射機(jī),自適應(yīng)接收機(jī)與目標(biāo)環(huán)境構(gòu)成了認(rèn)知雷達(dá)的動態(tài)系統(tǒng)結(jié)構(gòu)。本文首先對認(rèn)知雷達(dá)的知識輔助模塊進(jìn)行了研究,研究了海雜波的幅度分布模型和多普勒譜模型;接著將認(rèn)知雷達(dá)的體制與動目標(biāo)檢測(Moving Target Detecting,MTD)和空時自適應(yīng)處理(Space-time adaptive processing,STAP)相結(jié)合,針對認(rèn)知MTD和STAP雷達(dá)中發(fā)射和接收優(yōu)化問題,提出了認(rèn)知發(fā)射和認(rèn)知接收的優(yōu)化設(shè)計算法;最后通過仿真驗證,自適應(yīng)發(fā)射和自適應(yīng)接收能有效改善接收信號的信雜噪比(Signal to Clutter and Noise Ratio,SCNR),提高動目標(biāo)檢測性能。本文完成的主要工作如下:(1)詳細(xì)分析了海雜波的幅度分布模型和多普勒譜模型,利用IPIX雷達(dá)實測海雜波數(shù)據(jù)對幅度分布和多普勒特性進(jìn)行了分析,并且分別對低、高海況數(shù)據(jù)的幅度分布和多普勒譜模型進(jìn)行擬合仿真,作為后續(xù)認(rèn)知雷達(dá)動目標(biāo)檢測的雜波環(huán)境模型。(2)以SCNR為優(yōu)化設(shè)計準(zhǔn)則,建立了基于慢時間域脈沖編碼的MTD雷達(dá)信號模型,針對此模型進(jìn)行了一系列的簡化處理,得到錐約束下的二次型規(guī)劃(Cone Constrained Quadratic Programming,CCQP)問題;提出了基于簡約梯度法(Simplified Gradient Method,SGM)求解脈間編碼的算法;最后,與已有的重新設(shè)置參數(shù)的凸優(yōu)化算法(Convexification via Reparametrization,CoRe)和直接碼設(shè)計的循環(huán)算法(Cyclic Algorithm for Direct Code Design,CADCODE)進(jìn)行對比,計算機(jī)仿真結(jié)果表明:慢時間域編碼優(yōu)化能夠有效改善回波的SCNR為3dB左右,且該算法具有運(yùn)算量小和實時性能好的特點。(3)建立了認(rèn)知接收MTD濾波組的信號模型,并采用Lagrange對偶算法求解。將認(rèn)知接收MTD濾波組與已有的兩種自適應(yīng)MTD濾波器組做對比,仿真結(jié)果表明:認(rèn)知接收MTD濾波組具有凹口深,旁瓣低的特點;用IPIX真實海雜波數(shù)據(jù)來設(shè)計認(rèn)知接收MTD濾波器組,高于傳統(tǒng)加窗的FFT濾波器組輸出SCNR為10dB左右,對海雜波具有顯著的抑制效果。(4)建立了STAP脈間編碼信號模型,在選定目標(biāo)空時多普勒平面區(qū)間,提出了一種基于輔助變量和交替方向的方法結(jié)合的脈間編碼的優(yōu)化設(shè)計算法。實驗結(jié)果顯示了平行于雜波脊附近的檢測器輸出SCNR有一定的改善,有利于慢速目標(biāo)的檢測。同時建立了認(rèn)知STAP接收濾波器的模型,并采用Lagrange對偶算法求解。仿真結(jié)果表明:輸出信號的SCNR高于傳統(tǒng)自適應(yīng)濾波器的輸出,同時認(rèn)知接收濾波器的輸出旁瓣功率低于自適應(yīng)濾波器,旁瓣抑制性能較好。
[Abstract]:Cognitive radar (Cognitve Radar,CR) is a new radar system, which has the ability of learning, understanding and inferring environment, continuous information feedback from receiver to transmitter and waveform adaptive transmitting ability. Adaptive transmitter, adaptive receiver and target environment constitute the dynamic system structure of cognitive radar. In this paper, the knowledge assistant module of cognitive radar is studied firstly, the amplitude distribution model and Doppler spectrum model of sea clutter are studied, then the system of cognitive radar is combined with moving target detection (Moving Target Detecting,MTD) and space-time adaptive processing (Space-time adaptive processing,STAP). Aiming at the problem of transmitting and receiving optimization in cognitive MTD and STAP radar, an optimal design algorithm of cognitive transmitter and receiver is proposed. Adaptive transmission and adaptive reception can effectively improve the signal-to-noise ratio (Signal to Clutter and Noise Ratio,SCNR) of the received signal and improve the performance of moving target detection. The main work of this paper is as follows: (1) the amplitude distribution model and Doppler spectrum model of sea clutter are analyzed in detail. The amplitude distribution and Doppler characteristics are analyzed by using the measured sea clutter data of IPIX radar. The amplitude distribution of high sea condition data and Doppler spectrum model are simulated as the clutter environment model for moving target detection of cognitive radar. (2) based on SCNR, a MTD radar signal model based on slow time domain pulse coding is established. For this model, a series of simplified processing is carried out, and the quadratic programming (Cone Constrained Quadratic Programming,CCQP problem with conical constraints is obtained. An algorithm based on reduced gradient method (Simplified Gradient Method,SGM) is proposed to solve the inter-pulse coding problem. Compared with the existing convex optimization algorithm (Convexification via Reparametrization,CoRe) and the direct code design (Cyclic Algorithm for Direct Code Design,CADCODE) algorithm, the computer simulation results show that the slow time domain coding optimization can effectively improve the echo SCNR about 3dB. The algorithm has the advantages of small computation and good real-time performance. (3) the signal model of cognitive received MTD filter group is established and solved by Lagrange dual algorithm. Comparing the cognitive receiving MTD filter group with the two existing adaptive MTD filter banks, the simulation results show that the cognitive receiving MTD filter group has the characteristics of deep notch and low sidelobe, and the IPIX real sea clutter data is used to design the cognitive received MTD filter bank. Compared with the traditional windowed FFT filter bank, the output SCNR of the FFT filter bank is about 10dB, which has a remarkable suppression effect on the sea clutter. (4) the STAP inter-pulse coded signal model is established, and the selected target space-time Doppler plane is selected. An optimal design algorithm for interpulse coding based on the combination of auxiliary variables and alternate directions is proposed. The experimental results show that the output SCNR of the detector, which is parallel to the clutter ridge, is improved to some extent, which is beneficial to the detection of slow targets. At the same time, the model of cognitive STAP receiver filter is established and solved by Lagrange dual algorithm. The simulation results show that the output SCNR of the output signal is higher than that of the conventional adaptive filter, and the output sidelobe power of the cognitive receiver filter is lower than that of the adaptive filter, and the sidelobe suppression performance is better.
【學(xué)位授予單位】：南京航空航天大學(xué)
【學(xué)位級別】：碩士
【學(xué)位授予年份】：2017
【分類號】：TN957.51

【參考文獻(xiàn)】

相關(guān)期刊論文 前9條

1 史常麗;張勁東;邱曉燕;吳悅;;一種認(rèn)知MTD雷達(dá)脈間編碼設(shè)計方法[J];雷達(dá)科學(xué)與技術(shù);2016年04期

2 袁賽柏;金勝;朱天林;;認(rèn)知雷達(dá)技術(shù)與發(fā)展[J];現(xiàn)代雷達(dá);2016年01期

3 孫俊;;智能化認(rèn)知雷達(dá)中的關(guān)鍵技術(shù)[J];現(xiàn)代雷達(dá);2014年10期

4 黎湘;范梅梅;;認(rèn)知雷達(dá)及其關(guān)鍵技術(shù)研究進(jìn)展[J];電子學(xué)報;2012年09期

5 公緒華;孟華東;魏軼e，

本文編號：2253737