MIMO雷达自适应波形设计与阵列优化研究

发布时间：2018-07-05 02:35

本文选题：MIMO雷达 + 自适应波形设计　；参考：《哈尔滨工业大学》2016年博士论文

【摘要】：多输入多输出(Multiple input multiple output,MIMO)雷达已成为雷达信号处理领域中快速发展的一种新体制雷达,最近十几年受到了很多学者的关注。它的优势主要来自于其波形分集和灵活的阵列配置。随着对MIMO雷达研究的深入,发射波形不再仅限于选择正交波形,而是能够根据实际需求灵活设计发射方向图,将发射能量集中于感兴趣的区域或者逼近期望的方向图。这种由波形设计方向图的方法一般被分为两种:方向图匹配或最小旁瓣设计。方向图匹配一般不考虑峰值旁瓣过高的问题。同时,这种设计方向图的波形设计通常也忽略了其它自适应处理的要求,例如时域上抑制距离旁瓣和频域上抗有源干扰的需求。子阵级阵列与稀疏阵列是MIMO雷达常用的两种阵列配置。使用子阵级阵列的相控-MIMO雷达通常只考虑输出信噪比的改善,但却忽视了由于波形分集需要具备更高的抗干扰能力。MIMO雷达使用稀疏阵列可获取额外的自由度,但由于其方向图设计问题的复杂性而增加了阵列优化的难度,现有的算法很难满足这样的要求。本课题针对这些问题和缺陷,着重研究了MIMO雷达的自适应波形设计和阵列优化,并提供新的解决方法。首先,针对方向图匹配和抑制距离旁瓣干扰的两种需求,提出了迭代对角逼近算法(Iterative Diagonal Approximation Algorithm,IDAA)和任意维迭代谱逼近算法框架(Arbitrary-Dimensional Iterative Spectral Approximation Algorithm,ADISAA),分别解决了方向图匹配过程中旁瓣水平过高的问题和两种窄带波形的约束条件同时优化的问题。在IDAA算法中,方向图设计可被建模为期望方向图与导向矢量逆变换的卷积,通过使用迭代对角逼近的方式实现与期望方向图匹配的方向图。在ADISAA算法框架中,除了IDAA之外,本文通过对Multi-WECAN(Multi-sequence weighted cyclic algorithm-new)算法进行投影算子实现,使其能够作为其中一个子算法,由此设计指定区间相关旁瓣凹口抑制距离旁瓣干扰,进而实现了包含两种功能的窄带恒模波形优化。由此可见,将IDAA子算法与ADISAA结合之后,该算法框架可被视为一种高效的面向方向匹配的波形设计框架,可兼容其它约束条件的同时优化。仿真结果表明,相比现有的方法,ADISAA能够牺牲一部分的方向图匹配性能而获取更深相关旁瓣的凹口深度,且具有灵活的调整能力,更能满足MIMO雷达自适应波形设计的需求。其次,针对方向图匹配、抑制有源干扰和距离旁瓣干扰的三种需求,提出了一种多功能阻带波形设计方法,即使用ADISAA算法框架设计一种满足方向图匹配,指定区间良好的相关特性和功率谱凹口的多功能阻带波形。考虑到常规宽带波形的二维空间特性导致方向图旁瓣水平过高的问题,本文不对频率以及带宽进行严格的约束,而将其作为一个待优化变量,通过对秩亏傅里叶换算法进行投影算子的推导,将其作为ADISAA算法框架中的一个子算法,与方向图匹配和指定区间良好相关特性等投影算子结合,实现了一种面向方向图匹配的多功能阻带波形优化设计。第三,针对抗压制性干扰和频率正交信号优化的需求,提出了一种极化波束形成的方法和基于迭代矩阵谱逼近算法(Iterative Matrix Spectral Approximation Algorithm,IMSAA)的波形优化设计方法。本文考虑一种极化相控-MIMO雷达,即子阵单元为极化阵列,通过对零点凹口进行正交极化约束,有效地提高了对压制性干扰的抑制能力;接着,对子阵单元之间的频率分集信号的正交优化问题进行建模,使用IMSAA算法设计相关性能良好的发射信号。仿真验证了极化波束形成方法和波形优化算法的有效性。最后,针对MIMO雷达稀疏阵列优化设计的需求,提出了一种基于多目标差分进化(multi-objective differential evolution,MODE)算法的稀疏阵列优化方法,解决了方向图匹配和峰值旁瓣水平抑制的并行优化问题。在第一个步骤中,选择循环算法(cyclic algorithm,CA)设计满足满阵方向图匹配的协方差矩阵;在第二个步骤中,根据前面得到的协方差矩阵,对面向方向图匹配的稀疏阵列优化问题进行数学建模,引进遗传算法(Genetic algorithm,GA)和差分进化算法(Differential evolution,DE)求解该单目标优化问题。考虑到这两种单目标优化算法无法对峰值旁瓣水平进行抑制,将峰值旁瓣抑制作为一个不等约束条件,得到了一个约束优化问题,进而转为多目标优化问题。针对此,本文使用MODE算法求解该多目标优化问题。仿真结果表明:使用MODE算法能够设计满足方向图匹配和峰值旁瓣抑制的稀疏阵列。
[Abstract]:Multiple input multiple output (MIMO) radar has become a new system radar in the field of radar signal processing, which has been paid much attention by many scholars in the last decade. Its advantages are mainly from its waveform diversity and flexible array configuration. With the in-depth study of MIMO radar, the waveform is not transmitted. It is only to select the orthogonal waveforms, but to design the emission pattern flexibly according to the actual needs, to focus the emission energy on the region of interest or to approximate the expected direction. This method of pattern design by the waveform is generally divided into two kinds: the pattern matching or the minimum sidelobe design. The pattern matching generally does not consider the peak side. At the same time, the design of the design pattern is usually ignored by other adaptive processing requirements, such as the demand for the suppression of the distance sidelobe and the active interference in the frequency domain in the time domain. Subarray and sparse arrays are two commonly used array configurations for MIMO radar. A phased -MIMO radar using a subarray array is usually used. Only considering the improvement of the output signal to noise ratio, but ignoring the need for higher anti-interference ability of the waveform diversity,.MIMO radar can obtain extra freedom by using sparse array. But because of the complexity of the design problem of its directional map, the difficulty of array optimization is increased. The existing algorithm is difficult to meet such requirements. These problems and defects are focused on the study of adaptive waveform design and array optimization of MIMO radar, and provide new solutions. Firstly, the iterative diagonal approximation algorithm (Iterative Diagonal Approximation Algorithm, IDAA) and arbitrary dimensional iterative spectral approximation are proposed for the two requirements of the direction map matching and the suppression of the distance sidelobe interference. The Arbitrary-Dimensional Iterative Spectral Approximation Algorithm (ADISAA) solves the problem of high sidelobe level in the direction matching process and the simultaneous optimization of the constraints of two narrow band waveforms. In the IDAA algorithm, the pattern design can be modeled as the volume of the volume of the expected direction map and the direction vector inverse change. In the ADISAA algorithm framework, this paper implements the projection operator of the Multi-WECAN (Multi-sequence weighted cyclic algorithm-new) algorithm so that it can be used as one of the sub algorithms to design the specified interval phase in the framework of the Multi-WECAN (Multi-sequence weighted cyclic algorithm-new) algorithm. The close sidelobe restrains the distance sidelobe interference and realizes the optimization of the narrow band constant mode waveform containing two functions. This can be seen that, after combining the IDAA subalgorithm with ADISAA, the algorithm framework can be considered as a highly efficient direction oriented waveform design framework, which can be compatible with its constraint conditions at the same time. Compared with the existing methods, ADISAA can obtain deeper correlation sidelobe depth of concave mouth at the expense of a part of the pattern matching performance, and has flexible adjustment ability, and can meet the needs of adaptive waveform design of MIMO radar. Secondly, three requirements for active interference and distance sidelobe interference are suppressed. The design method of the functional stopband waveform is to use the ADISAA algorithm framework to design a multi-function band waveform that satisfies the direction map matching, specifies the good correlation characteristics of the interval and the power spectrum concave. Considering the problem of the high level of the sidelobe in the direction map, the two-dimensional space characteristics of the conventional wide-band waveform will not be strict with the frequency and bandwidth. The constraint of lattice is used as a variable to be optimized. By deriving the projection operator from the rank deficiency Fourier transform algorithm, it is used as a sub algorithm in the framework of ADISAA algorithm. It combines with the projection operator, which is matched with the direction map and the good correlation characteristic of the specified interval, and realizes a multi-function band waveform optimization for the directional map matching. Third, in view of the demand for compression interference and frequency orthogonal signal optimization, a method of polarization beam formation and an optimization design method based on Iterative Matrix Spectral Approximation Algorithm (IMSAA) are proposed. A polarization phased -MIMO radar, the subarray unit, is considered in this paper. For the polarization array, the suppression ability to suppressing interference is effectively improved by orthogonal polarization constraint on the zero point notch. Then, the orthogonal optimization problem of the frequency diversity signal between the subarray units is modeled and the IMSAA algorithm is used to design the good correlation signal. The polarization beam forming method and wave are verified by simulation. In the end, a sparse array optimization method based on multi target differential evolution (multi-objective differential evolution, MODE) algorithm is proposed for MIMO radar sparse array optimization design, which solves the parallel optimization problem of directional map matching and peak side lobe suppression. The first step is the first step. In this case, cyclic algorithm (CA) is designed to satisfy the covariance matrix matching the full array direction map. In the second steps, the mathematical modeling of the sparse array optimization problem matching the opposite direction map is modeled according to the covariance matrix obtained before, and the genetic algorithm (Genetic algorithm, GA) and the differential evolution algorithm (Differential) are introduced. Evolution, DE) solves the single objective optimization problem. Considering that the two single objective optimization algorithms can not suppress the peak sidelobe level, the peak sidelobe suppression is regarded as a unequal constraint condition, and a constrained optimization problem is obtained, and then to the multi-objective optimization problem, the MODE algorithm is used to solve the multi-objective optimization problem. Simulation results show that the MODE algorithm can be used to design sparse arrays satisfying pattern matching and peak sidelobe suppression.
【学位授予单位】：哈尔滨工业大学
【学位级别】：博士
【学位授予年份】：2016
【分类号】：TN957.51

【相似文献】