【作者】 初萍；

【导师】 司伟建；

【作者基本信息】 哈尔滨工程大学 ， 通信与信息系统， 2011， 博士

【摘要】 围绕实际工程应用中被动测向技术面临的问题展开研究,主要对任意阵列的干涉仪测向方法、提高MUSIC (Multiple Signal Classification)算法的分辨力、实际测向系统通道的幅相不一致校正方法几个方面进行了深入的研究,提出了相应的解决方法,并进行了软硬件方案的设计和调试,在实际测向系统中对提出的方法进行了相应的测试,验证了提出方法的可行性。基于干涉仪测向方法提出了立体基线测向方法,它可以在任意的天线阵列形式下实现宽频带内辐射信号入射角度的无模糊求解。建立了立体基线测向模型,对任意两个阵元接收数据的相位差与信号入射角度的关系进行了推导,通过建立方程组实现对信号入射角度的计算,并提出了多值模糊和镜像模糊的解决方法,通过仿真实验分析了立体基线测向方法对入射信号角度的估计性能,并且通过实际测向系统对立体基线测向方法进行了测试。针对阵列孔径限制分辨力及基于四阶累积量的算法计算量大等问题,提出了改进的四阶累积量二维MUSIC算法,该算法根据分辨力与阵列孔径成正比的特点,选择构成最大虚拟孔径的一组阵元进行二维入射角度估计,同时给出了新的阵列流型及四阶累积量矩阵的构造方法。最大的虚拟阵列孔径是真实阵列孔径的2倍,且其阵元数与真实阵元数相同,获得虚拟阵列对应的四阶累积量矩阵后根据MUSIC算法步骤进行角度估计,可以提高有限阵列孔径条件下MUSIC算法的分辨力且基本不增加计算量,并且通过仿真实验验证了理论的有效性。在配有诱饵的雷达系统中,为了实现从各个入射角度都能充分的保护雷达信号,诱饵脉冲前沿通常会超前雷达脉冲信号,结合诱饵和雷达信号的辐射特点提出了基于不完全重合信号的MUSIC算法,利用多个信号的协方差矩阵与单个信号的协方差矩阵之间的关系,可以得到对应入射信号的协方差矩阵,通过对各个协方差矩阵的处理可以依次估计得到各个信号的入射角度,仿真及实际系统接收数据的测试均验证了提出算法的可行性。针对空间谱估计算法性能对阵列通道幅相不一致误差敏感的问题,根据实际测向系统接收通道的特点,提出了利用自检信号进行实时校正的方法,将整个通道的校正分为天线到接收通道之间的无源器件预校正和接收通道的实时校正两个部分,在测向过程中利用自检信号对接收通道进行实时校正。仿真结果表明实时校正方法可以动态校正通道的幅相不一致误差并且有利于提高MUSIC算法的分辨力,实测数据也表明实时校正方法可以动态的校正通道幅相不一致误差,达到提高测向精度的目的。对测向信号处理器进行了方案设计、硬件制作及整机测试,介绍了测向信号处理器的硬件组成及其软件程序流程图,在实际测向系统中对立体基线测向方法和MUSIC算法的运算时间以及测向精度进行了测试,技术指标满足项目要求。更多还原

【Abstract】 It is about the problem that passive direction finding techniques faced in actual engineering application. It mainly studies on such aspects of interferometer direction finding method of arbitrary array, improving resolution performance of MUSIC (Multiple Signal Classification) algorithm and calibration method of amplitude-phase mismatch in actual direction finding system, and corresponding methods are proposed. The software and hardware of direction estimation processor is designed and debugging at the same time. The proposed methods are tested in actual direction finding system, and their feasibility is proved adequately.Direction finding method of spatial baseline is proposed based on interferometer method. It can realize unambiguity estimation of directon of arrival with arbitrary arrays in wide band. Its direction finding model is established and the relationship between phase difference of two arbitrary antennas and direction of arrival is derived. It realizes estimating the direction of arrival by constrrcting equations. Methods of resolving ambiguity of multi-value and mirror are proposed in the process of computing.To the problems of resolution limited by array aperture and large computation cost of algorithms based on fourth-order cumulant, a modified 2-D MUSIC algorithm based on fourth-order cumulant is proposed. According to the theorem that resolution is proportional to array aperture, a group of antennas which constitute the largest virtual aperture is chosen for 2-D direction of arrival estimation, and the new array manifold and construction method of fourth-order cumulant matrix are proposed. The largest virtual array aperture is twice of the actual ones, and its antenna number is the same to the actual ones. According to the new array manifold and fourth-order cumulant matrix, we can realize 2-D direction of arrival estimation following steps of MUSIC algorithm. It prompts resolution validly with limited array aperture and no computation cost in addition. Its availability is testified by simulation experiment.In radar system which is equipped with decoy, usually, the decoy pulse is advanced by radars to protect the radar adequately. According to the receiving characteristic of pulse signals in actual direction finding system, MUSIC algorithm based on some overlap signals is proposed. It can obtain a new covariance matrix of signal by relationships between covariance matrixes of multi-signals and single signal. Then, it can estimate directions of arrival by processing each covariance matrix successively. It testified the feasibility of the proposed algorithm by simulation results and test in actual direction finding system. To solve the problem that the property of spatial spectrum estimation algorithms is sensitive to array channel amplitude-phase mismatch, a real-time calibration method based on self-checking signal is proposed according to the characteristic of actual receiving channel. The whole channel calibration is divided into two parts:the pre-calibration of passive devices between antennas and receiving channels and the real-time calibration of receiving channels. At first, it can get the pre-calibration matrix by setting a signal which has direction of arrival of 0°. Then, it can realize real-time calibration using self-checking signal during the process of direction finding. It is proved by simulation results that the proposed real-time calibration method can dynamically calibrate the channel amplitude-phase mismatch errors and it is helpful to prompt the solve ability of MUSIC algorithm. The experiment in actual direction finding system has also testified that it can calibrate amplitude-phase mismatch real-timely and get higher direction finding precision.According to technical intexes, proposal design, hardware manufacture and test are done to direction finding signal processor. The structure of hardware is introduced in detail, and its software flow chart is also introduced. The computing time of spatial baseline and MUSIC algorithm are test in actual direction finding system. Also, their direction finding accuracy are test in direction finding system. The results of testing in actual system prove that the processor achieve the technical intexes well.更多还原