【作者】 杨博；

【导师】 周一宇；

【作者基本信息】 国防科学技术大学 ， 信息与通信工程， 2011， 博士

【摘要】 利用空基辐射源照射的非合作探测系统能够实现对空中或地面目标的探测。它具有结构简单、成本较低、探测范围广、战场隐蔽性高及对隐身目标的探测能力等诸多优点,但是非合作空基辐射源的位置、速度、天线波束指向、脉冲发射时间等因素均未知且不可控制,使得该系统在具有诸多优势的同时存在着一系列需要解决的技术问题。本文从系统设计与分析、杂波建模、目标信号检测、多目标定位与数据关联等方面开展了非合作探测系统的理论和仿真研究,旨在从理论上解决该背景下无源探测系统所面临的主要问题,为非合作无源探测系统的发展与完善提供一定参考和借鉴。论文第二章在分析空基辐射源特点的基础上,选择预警机雷达信号作为外辐射源信号,并且以某一型号预警机为例进行了具体讨论,给出了系统设计思路和基本结构,通过理论分析来预测系统所能达到的主要技术指标,阐明系统的实用性和可行性。针对空基辐射源波束照地形成的强地杂波现象建立了地杂波的数学模型,分析了地杂波的空-时-频特性,分析结果表明地杂波与目标回波信号在时-频域不可分,但在更高维的空-时-频域上可分,为杂波环境下的目标检测奠定了基础。论文第三章分析了非合作目标检测中长时间相干积累带来的“距离门走动”问题,给出了基于参考信号距离拉伸的时延补偿方法,仿真结果表明该方法能够有效地解决相干积累时间内的时延补偿问题,有效地聚集多个脉冲回波信号的能量,提高输出信噪比。针对地杂波对运动目标检测的影响,提出了基于混合积滤波的非合作目标检测方法。该方法利用杂波与目标在空-时-频域的可分性,采用阵列天线接收方式采集目标及杂波的方位信息,借鉴双基地STAP方法的基本思路,将非合作领域中一般的时-频检测方法推广为时域混合积数据的空域-频域二维联合滤波,仿真结果表明当目标信号与杂波信号在空-时-频域不出现重合时,该方法能够有效地抑制杂波和接收机噪声,实现地杂波背景下的动目标检测。另外,基于混合积滤波的非合作目标检测方法中存在着双基地STAP距离独立性补偿的难点。针对现有的基于映射的距离独立性补偿方法中存在的不足,提出了基于变换域的距离独立性补偿方法,克服了原有方法中处理过程复杂、引入误差项过多的缺点,提高了估计的准确性。针对中、高重频条件下的距离模糊现象,提出了基于数据分离的距离独立性补偿方法,该方法通过将数据在频域上先分离后合并的手段来实现存在多个模糊距离门时的杂波协方差矩阵估计,仿真结果表明该方法能够有效地改善存在距离模糊时的估计效果。论文第四章在分析静态阈值随机共振和双稳态随机共振现象的基础上,讨论了随机共振方法在非合作目标检测中的应用。分别从非周期信号随机共振和周期信号随机共振两个角度给出了相应的非合作目标检测方法,并对两种方法进行了比较。基于非周期信号随机共振和基于周期信号随机共振的非合作目标检测方法将非合作检测与非线性模型相结合,利用非线性系统中的随机共振现象,通过系统、输入信号及噪声的协同来增强输出信噪比,达到提高目标检测性能的目的。另外,在静态阈值模型的基础上进一步提出了一种渐进最优的阵列静态阈值随机共振模型,该模型利用多个静态阈值模型的组合以及多参数的调节到达提高弱目标信号检测性能的目的。论文第五章对空基非合作照射中系统的目标定位问题进行了研究。根据空基非合作照射的结构特点分别提出两种目标定位方案:基于运动辐射源的三站定位方案和四站定位方案,给出每种方案下对应的多目标数据关联方法。时差定位精度的仿真结果表明两种定位方案都能够实现对主要观测区域的高精度定位,其中三站定位方案的定位精度分布会受到辐射源位置的影响,而四站定位方案的定位精度分布则主要取决于地面站的位置。为了进一步提高定位精度,提出了一种基于约束总体最小二乘的定位解算方法,该方法由于考虑了误差因素对系数矩阵的影响,并对此进行了最小化处理使得解算后的误差要小于直接由解析方法求解产生的误差。更多还原

【Abstract】 Non-cooperative bistatic radar systems based on Spacial Motorial Emitter can detect wide distance targets in the ground or in the air. They have the advantages that the receivers are potentially simple and cheap. Bistatic radar may have a counter-stealth capability, since target shaping to reduce monostatic RCS (Radar Cross Section) will in general not reduce the bistatic RCS. In spite of those advantages, they have to face much difficulty such as unknown Emitter’s position and velocity and antenna beam direction. All of these make the target detection become hardness and should be solved by corresponding technology approachs. This dissertation includes system design, clutter analyses, target detection, target location and data association. It aims to overcome the interrelated technology difficulties on this spacial non-cooperative background. All of research content may be the reference for the development of Non-cooperative bistatic radar systems.In Chapter 2, AEW (air early warnning) radar signal is chosen to be the emitter. The system design and framework are present on this background. Its performance is analyzed base on certain radar parameters, results reveal the practicability and feasibility of system. At the same time, ground clutter’s space-time-frequency characteristic is studied; the result shows that it is not divisible between ground clutter and targets in the time-frequency fields, while they are divisible in the space-time-frequency union fields.In Chapter 3, the algorithm of long-time coherent integration is researched when the range migrations exist in the echo signal. A method of drawing or pulling reference signal is proposed, simulation result shows the method can assemble multi-pulses energy and improve signal-noise-ratio. In order to realizing target detection in clutter and noise background, a target detection algorithm based on mixing product filtering is pointed; it spreads the traditional time-frequency detection to the space-time-frequency union filter. Simulation result shows the effectiveness of the method on the conditional of target signal without covering by clutter in space-time-frequency union fields. In addition, mixing product filtering target detection algorithm relates to a difficulty about range-dependence compensation for bistatic STAP. Because of the RBRDC’s (Registration Based Range Dependence Compensation) defect, a method named range-dependence compensation method based on transformation (RBTRC) is proposed. RBTRC omits needless transformation process, reduces the estimate errors and improves operation efficiency. Aim at the existing compensation method’s deficiency on the condition of range ambiguity, a new method named range-dependence compensation based on data separation was proposed. Data simulation shows that the new method has the better signal-to-noise upgrading performance for range ambiguity radar signal.In Chapter 4, the application of stochastic resonance (SR) theory in non-cooperative signal detection field is discussed. Fistly, static threshold system and dynamical bistable system’s SR phenomena are studied; secondly non-cooperative target detection methods using periodic stochastic resonance or aperiodic stochastic resonance are proposed. Unlike traditional non-cooperative detection, SR can achieve high-point output SNR by harmonizing the nonlinear system, signal and noise.In Chapter 5, the target location about non-cooperative bistatic radar systems is researched. Two location schemes using TDOA (time difference of arrival) are proposed; one is composed of three receiving stations and emitter, the other is composed of four receiving stations. Their multi-target measurement data association methods are also offered. The simulation result shows target location precision is affected by the position of the emitter in the first method, while target location precision is independent of the position of the emitter in the second method. In addition, a restricted total least squares (RTLS) algorithm based on TDOA is applied to multi-station passive location in order to reduce the target position errors. This algorithm takes into account coefficient matrix errors in the process of calculations, so it has better performance than analytical arithmetic.更多还原