【作者】 孙学军；

【导师】 唐斌；

【作者基本信息】 电子科技大学 ， 信息获取与探测技术， 2010， 博士

【摘要】 阵列信号处理经过几十年的发展,已经取得了诸多研究成果。但在一些工程具体应用中仍有较多关键问题亟待解决。如天线阵列拓扑设计、微弱信号测向、波束形成、互耦和通道误差校正等方面,在工程应用中的实现研究仍是难点。特别是利用星载平台对地面弱小信号目标进行侦测,面临着阵列单元数受限、可供布阵的区域形状和面积受限、最小单元半径受限、接收信号微弱、互耦和通道误差校正难度较大等不利条件,需要对星载阵列天线的阵列拓扑结构进行优化设计,结合阵列拓扑结构对微弱多信号测向和波束形成技术进行研究,对互耦和通道误差进行校正,以提高接收灵敏度和测向精度。论文主要针对星载阵列天线的布阵方式及优化方法、阵列测向方法、数字波束形成技术以及阵列误差校正方法等关键技术进行研究。文中提出、推导了一系列具有理论及实用价值的新算法,并通过计算机仿真实验验证了所提算法的优良性能。归纳起来,本文的贡献和创新主要在以下几个方面:1、基于微遗传优化算法和粒子群优化算法方法,研究了阵列拓扑结构设计及优化方法,分别设计了一维非均匀线阵、七臂圆环阵列和修改的七臂圆环阵列结构。对一维非均匀线阵、七臂圆环阵列和修改的七臂圆环进行了综合分析和仿真,给出了综合及仿真结果。仿真表明,在给定的条件下,七臂圆环阵列在副瓣性能上具有更好的特性。2、提出了基于二次组阵的MUSIC测向算法,该算法在阵列中选择多个子阵构成的子阵集合,可以通过子阵间阵元的交叉或重叠选取的方式增加整个阵列天线的阵元增益,从而实现对微弱信号的DOA估计。分别结合七臂圆环阵和修改的七臂圆环阵对测向性能进行了仿真,仿真验证了该方法的有效性。提出了基于二次虚拟内插技术的圆阵接收2D-DOA分离估计方法,该方法通过对圆阵进行了两次内插,根据内插出的两个圆阵之间的移不变性,可分别利用ESPRIT算法和MUSIC算法直接获得俯仰角和方位角的估计。仿真研究表明二次内插比单次内插的估计精度高。3、研究了基于协方差对角加载、广义线性组合和幅度加权的波束形成技术,设计了幅度加权方式,结合七臂圆环阵对这些波束形成方法进行了仿真,仿真表明采用圆环幅度加权形式,能够在较宽的频率范围内降低旁瓣。提出了基于粒子群优化的波束形成方法,该方法易于并行实现,能极大地提高算法的运算速度。经仿真验证,算法与SMI算法性能相当,即使存在阵列通道误差时,该算法也能有效收敛。研究了并行多波束形成方法和基于线性约束多波束形成方法,结合七臂圆环阵进行了仿真,仿真验证了这两种方法在七臂圆环阵情况下可有效地形成多波束,波束形成效果好,能同时分离多个信号。4、针对存在互耦情况下的阵列输出模型与阵列校正问题,提出了一种新的阵列互耦有源校正算法。该算法基于均匀圆阵,适用于校正多个源同时存在的情况。更多还原

【Abstract】 After decades of advance in array signal processing, there are numerous research achievements. However in realistic application, there are still many key problems are to be solved. For example, antenna array structure design, direction finding for weak signal, beamforming, calibration of mutual coupling and channel errors, these problems are still hard to be solved. Especially when using satellite platform to sense the weak signal on the ground, we are facing many problems such as limitation of the number of array units, limitation of the area and the shape of the platform for array disposing, limitation of the radius of the minimum unit, weakness of the received signal, the difficulty of emendation of mutual coupling and channel errors. We have to optimize the design of the array structure for satellite carrier, research the direction finding and beamforming method for weak signal based on the array structure, amend the mutual coupling and channel error according to the impact of the direction finding and beamforming, thus we can improve the receive sensitivity and the precision of direction finding. In this dissertation, several key techniques in array signal processing such as the collocation and optimization method of the antenna array, direction finding method, digital beamforming and array calibration method are to be studied. Several novel and valuable algorithms were proposed and derived in the dissertation, computer simulation validated their efficientness. The main contributions and innovations of this dissertation are as follow:1. Based on the micro-genetic optimization algorithm and the particle swarm optimization algorithm, the array collocation and optimization methods are studied, one dimensional nonuniform linear array, seven arm circle array and modified seven arm circle array are designed. We synthesized, analyzed and simulated the performance of the one dimensional nonuniform linear array, seven arm circle array and modified seven arm circle array, the synthesized and simulation results are given. The simulations show that the seven arm circle array owns the lowest sidelobe level in the same condition.2. The MUSIC algorithm based on twice organized array is proposed, this algorithm chooses multiple subarray in the array as subarray set, thus improving the array gain for the whole array by crossing or overlapping the subarrays, and realize the direction of arrival(DOA) estimation for weak signal. Combining with seven arm circle array and modified seven arm circle array, we simulated the the direction finding algorithm, the simulation validated the efficiency of the algorithm. The circular array received 2D-DOA separable estimation method based on twice virtual interpolation is proposed. This method interpolated the circular array twice, according to the shift invariance of the two interpolated virtural circular array, we can obtain the estimation of the azimuth and elevation by using ESPRIT and MUSIC algorithm respectively. Computer simulation shows that the performance of twice interpolation is superior than single interpolation.3. Single beam forming methods which based on covariance matrix load, amplitude weighted using generalized linear combination are studied. The amplitude weight method is designed. We simulated these beamforming algorithms based on seven arm circle array, simulation results shows that we can get lower sidelobe in wider band by using circle amplitude weighted method. Beamforming method based on particle swarm optimization algorithm is proposed. This method is easy to realize, can improve the computation efficiency greatly. From the simulation, we can see that the performance of this algorithm is as good as SMI algorithm, when there is channel error, the algorithm can also be convergence efficiently. Parallel and linearity restriction multiple beamforming method were studied respectively. The simulation is based on seven arm circle array, the simulation validated the efficiency of the two algorithm, they can beamform efficiently, and can separate multiple signals.4. The array output model and array calibration problem are studied in presence of mutural coupling, a new active array calibration algorithm is proposed to solve the problem of mutural coupling. This algorithm is based on uniform circular array. It is suitable to calibrate the array when multiple sources exist.更多还原