基于中小孔径阵列的超分辨测向算法

发布时间：2018-05-18 15:45

本文选题：超分辨测向算法 + 被动测向　；参考：《哈尔滨工业大学》2015年博士论文

【摘要】：随着移动通信、遥测遥感与航空航天等技术的发展,依靠目标辐射的电磁信号进行被动方位检测的技术越来越受到重视,能够搭载在舰船、车辆、飞机、卫星等移动平台上的中小型阵列,成为被动测向技术的重要应用平台。近年来高频与微波等电子元器件的小型化使得整套天线、微波以及后端处理系统均可集成到很小的空间中,仅占用很小的体积和载荷。但和强大的硬件性能不相匹配,实际使用的被动测向算法测向精度有限,而且难以有效分辨多个空间邻近信号。近几十年,以MUSIC与ESPRIT算法为代表的子空间类算法得到广泛研究,这类算法具备接近克拉美罗界的优良测向性能和较强的多信号分辨能力,尤其适用于中小型阵列被动测向系统。但由于工程应用中遇到的问题,并未得以广泛采用。本文针对MUSIC算法在实际系统中遇到的几种重要问题展开研究,给出了相应的解决方案,并自主设计和实现了一套能够实际运行MUSIC及其扩展算法的小口径阵列试验系统。实际环境中存在大量窄带极化信号。利用信号的极化信息提升算法的测向精度与分辨力性能,对被动测向系统有重要意义。文中对常规MUSIC算法进行了扩展,使其能够应用于极化敏感阵列,并对扩展后MUSIC算法的空间谱性质进行了分析,给出一种能够显著降低算法运算量的快速极化MUSIC算法,并对极化敏感阵列在算法精度与分辨力性能上带来的提升进行了理论分析。通过蒙特卡洛仿真说明了算法在精度与分辨力性能上的优越性,并给出了计算复杂度的统计结果。理论和仿真结果均说明了算法的有效性。虽然MUSIC等子空间算法具备优良的测向与分辨能力,但仍需考虑阵列各类特性对算法的影响,进行合理的阵列设计以充分发挥算法性能。阵列的特性主要包括阵元的排列方式及其方向图特性。与传统DBF等测向算法相似,MUSIC等子空间算法也存在阵列模糊等问题。仅依靠提高阵元密度来克服阵列模糊,会导致算法的复杂度过高,而较大的阵列口径会使需要密集的空间谱采样计算,提高计算复杂度。文中设计了一种由不同口径阵列合成的组合阵列,利用阵元密集的小口径阵列避免阵列模糊现象,阵元稀疏的大口径阵列提高算法的性能,并对MUSIC算法做出一定修改,使得算法的运算量得以降低。在阵元方向图特性方面,本文提出利用阵元方向图中的幅度和相位的不一致特性提升算法性能的构想并设计了一种考虑阵元特性的修正MUSIC算法,对该构想进行了理论上的验证。通过蒙特卡洛仿真进一步证实了该方法的可行性。MUSIC等子空间类算法普遍存在难以识别多个相干信号的弱点。实际环境中,多径、有源雷达诱饵等因素都会导致相干信号的出现。有效估计相干信号的入射方向,成为子空间算法在实际应用中亟待解决的问题。考虑到中小型阵列通常架设于移动平台上,并受到空间前后向平滑解相干算法的启发,文中研究了一种能够处理相干信号的方案,并针对平台移动的速度给出不同方法。对高速平台和曲线运动平台的解相干算法进行了较详细的理论分析,在已知平台的移动特性、阵列的口径和信噪比、快拍数等因素时,可据此计算阵列能否处理相干信号。文中给出了仿真结果,说明算法的解相干效果。为证实MUSIC等子空间算法在中小型阵列被动测向系统中的可行性,我们搭建了一个运行快速极化MUSIC算法的被动测向实验系统。系统采用半径不超过1倍波长的小型均匀极化敏感圆阵,对较低频率范围的信号进行数据采集和方位信息估计。实验取得了较为理想的信号方位估计与分辨结果。
[Abstract]:With the development of mobile communication, remote sensing and Aeronautics and Astronautics, the technology of passive azimuth detection based on electromagnetic signals from target radiation is becoming more and more important. It can be carried on small and medium array on mobile platforms such as ships, vehicles, aircraft, satellites and so on. It has become an important application platform for dynamic direction finding technology. The miniaturization of wave and other electronic components makes the entire antenna, microwave and back-end processing system can be integrated into small space, taking only a very small volume and load. But it is not matched with the powerful hardware performance. The precision of the passive direction finding algorithm used is limited, and it is difficult to effectively distinguish a number of space adjacent signals. In the past ten years, the subspace class algorithm, represented by MUSIC and ESPRIT algorithm, has been widely studied. This algorithm has excellent direction finding performance and strong multi signal resolution near the crameno boundary. It is especially suitable for small and medium array passive direction finding systems. However, the problems encountered in engineering applications are not widely used. Several important problems encountered in the actual system are studied by MUSIC algorithm, and the corresponding solutions are given. A set of small aperture array test system which can actually run MUSIC and its extended algorithm is designed and implemented independently. There are a large number of narrow band polarization signals in the actual environment. The performance of precision and resolution is of great significance to the passive direction finding system. In this paper, the conventional MUSIC algorithm is extended to be applied to the polarization sensitive array, and the spatial spectrum properties of the extended MUSIC algorithm are analyzed. A fast polarization MUSIC algorithm which can significantly reduce the computation of the algorithm is given, and the polarization sensitive array is also given. The improvement of algorithm precision and resolution performance is analyzed. The superiority of the algorithm in accuracy and resolution performance is illustrated by Monte Carlo simulation, and the statistical results of computational complexity are given. Both the theoretical and simulation results show the effectiveness of the algorithm. Although the MUSIC subspace algorithm has excellent performance, the algorithm is excellent. The ability to measure direction and resolution, but still need to consider the influence of various characteristics of array on the algorithm, and carry out a reasonable array design to give full play to the performance of the algorithm. The characteristics of the array mainly include the arrangement of the array element and the characteristics of its direction map. Similar to the traditional DBF and other algorithms such as the direction finding algorithm, the MUSIC sub space algorithm also has the array ambiguity and so on. High array density to overcome array ambiguity will lead to high complexity of the algorithm, and large array aperture will require intensive spatial spectrum sampling and computational complexity. A combination array composed of different aperture arrays is designed to avoid array ambiguity and sparse array element. The large aperture array improves the performance of the algorithm and makes some modifications to the MUSIC algorithm, which reduces the computational complexity of the algorithm. In the aspect of the character of the array element direction map, this paper proposes a concept of improving the performance of the algorithm by using the inconsistency of the amplitude and phase in the array element direction graph, and sets up a modified MUSIC algorithm considering the characteristic of the array element. Through Monte Carlo simulation, the feasibility of the.MUSIC subspace class algorithm is proved to be difficult to identify multiple coherent signals. In the actual environment, the multipath, active radar decoy and other factors will lead to the emergence of the phase dry signal. Shooting direction is a problem to be solved in practical application. Considering that small and medium array is usually erected on mobile platform and inspired by the algorithm of smoothing the coherent signal before and after space, a scheme to deal with coherent signals is studied in this paper, and different methods are given for the speed of platform movement. The decoherence algorithm of the platform and the motion platform of the curve is analyzed in detail. In the case of the known platform's moving characteristics, the array aperture, the signal to noise ratio, the fast beat number and so on, the array can be calculated if the coherent signal can be processed. The simulation results are given to illustrate the algorithm's decoherence effect. In order to verify the MUSIC subspace algorithm, In the small and medium array passive direction finding system, we have built a passive direction finding experiment system with fast polarization MUSIC algorithm. The system uses a small uniform polarization sensitive circular array with a radius of not more than 1 times the wavelength, and the data acquisition and azimuth information estimation of the signal in the lower frequency range. The experiment is ideal. The estimation of the azimuth and the resolution of the signal.
【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】：2015
【分类号】：TN911.7

【相似文献】