【作者】 姚静；

【导师】 易东云；

【作者基本信息】 国防科学技术大学 ， 系统分析与集成， 2008， 博士

【摘要】 分布式SAR卫星系统采取小卫星编队飞行搭载SAR的方式,通过编队卫星与星载SAR协同工作,可实现传统单星SAR无法完成的多项测绘任务,是一种具有巨大潜力的新概念新体制雷达系统,其实现在基础理论和技术层面上面临许多挑战。其中首要问题之一就是确定卫星系统的空间状态。空间状态确定方法取决于状态测量方案、小型航天器设计等因素,服务于有效载荷任务、编队协同控制等方面。获取高精度卫星空间状态是分布式SAR任务成功实现的重要保障。出于合理有效利用未来空间资源的思想,本文探索性地开展了将多频率多星座GNSS观测数据用于分布式SAR卫星系统空间状态确定的研究。本文的研究工作和贡献主要体现在以下几个方面:第一,在明确分布式SAR卫星系统空间状态概念与模型的基础上,给出了精度需求分析与参数内在联系。通过需求分析说明了空间状态特别是相对位置确定在整个分布式SAR卫星系统中的重要地位,明确了空间状态与InSAR干涉基线在系统中的层次;分析空间状态自身特征以给出提高确定精度的思路,对分类中的绝对状态与相对状态之间的相互作用关系加以关联建模与误差分析,从而为分布式SAR卫星系统的顶层设计提供参考。第二,完整建立了分布式InSAR卫星系统测量基线与干涉基线的关联,实现了二者的转换与精度分析。细致分解了由测量基线获取干涉基线的全过程,将空间状态中除相对位置之外的其它参数作为影响转换流程的边界条件,明确了转换中的参数信息流程,进行了精度分析。针对非自主式测量模式,兼顾自主式测量模式,建立了从测量基线到干涉基线的空间域转换模型,对误差传播关系进行了全面系统的研究;针对测量数据采样率的不足,研究了从低数据率基线到高数据率基线的时间域转换方法,从函数逼近角度分析了基线的保精度高数据率插值方法中的逼近误差和随机误差综合影响,并考虑配准偏移量的影响,完善了由测量基线到干涉基线的转换流程。第三,建立了基于多频GNSS星间相对定位的节省参数样条表示模型,并提出了适合编队环境的多频GNSS整周模糊度解算方法。一方面,综合利用观测数据与待估参数的几何关系以及待估参数随历元连续变化的约束,基于函数逼近和节省参数建模理论,提出了基于多频GNSS的相对定位样条表示模型,展示了多频数据相对于双频数据、样条表示模型相对于传统逐点模型在改善星间相对位置解算精度方面的理论基础与仿真实例。另一方面,研究了GNSS相对定位中的模糊度解算这一关键问题,基于几何无关或相关模型、原始或组合观测值、搜索或序贯取整等不同条件所带来的性能差异,给出基于几何相关和整数降相关变换后取整的层叠式三频模糊度固定方法,具有在编队环境下单历元固定模糊度的性能和较好的抗差能力。第四,进行了基于组合GNSS的卫星空间状态确定性能评价,提出了面向分布式InSAR性能的导航卫星优选准则。定性给出了组合GNSS共用的优势;利用抗差性、载波相位整周模糊度解算成功率特别是观测几何等定量评价指标,给出了组合GNSS模式带来空间状态确定性能提高的全面仿真算例和严格理论依据。综合考虑观测数据冗余及其贡献,联合GNSS空间状态测量子系统和InSAR测高子系统,从面向高程精度重建的角度提出了新的导航卫星优选准则,在相同观测条件下可提升分布式SAR的性能。更多还原

【Abstract】 The distributed SAR satellite system is realized by fixing the synthetic aperture radars on formation flying small satellites. Through the collaboration of the formation flying satellites and the spaceborne SAR, the system is capable of performing various surveying and mapping tasks which cannot be imagined under the conventional single satellite SAR mode. It is a kind of new conceptual radar systems with enormous potential. There are many challenges in both the basic theories and the technical level before it comes into reality. One of the most important issues is the determination of the spatial states of the satellite system. The determination method depends on several factors, such as the states measurement schemes and the small spacecraft design. It also serves the payload missions, the formation cooperative control, and so on. The acquisition of high-precision satellite spatial states is the important guarantee for the success of distributed SAR missions. Based on the idea of utilizing the future space resources properly and effectively, this dissertation tries to probe into the spatial states determination method of distributed SAR satellite system based on the multi-frequency and multi-constellation GNSS observations.The researches and contributions are mainly embodied in the following areas.Firstly, the precision requirement analysis and internal relationships of parameters are presented based on the concepts and models arrangements of the distributed SAR satellite system spatial states. The significances of the spatial states determination, especially the relative position determination are illustrated through requirement analysis of the entire distributed SAR satellite system. The levels of the spatial states and the InSAR interferometric baseline are defined in the system. The characteristics of the spatial states are analyzed to study the clues of accuracy improvement. The interactions between the absolute and relative states in the classifications are associated with modeling and error analysis. The work can offer references for the top-level design of the distributed SAR satellite system.Secondly, the whole relationship is established between the measurement baseline and the interferometric baseline of the distributed InSAR satellite system. The conversion and precision analysis are also carried out. The entire process from the measurement baseline to the interferometric baseline is decomposed in detail. The parameters among the spatial states, except the relative position, function as boundary conditions of the conversion. The parameters information flow and precision analysis are put forward. The conversion models in the space domain are established from the measurement baseline to the interferometric baseline. The non-autonomous measurement mode with the autonomous one is taken into account in the modeling, and comprehensive studies are performed on error propagation relationships. Due to the reality of the low measurement sampling rate, the conversion models in the time domain are established from the low rated spatial states to the high one. The precision-keeping and high-rate interpolation method for the baseline is raised and the combined influences of approximation errors and random errors are analyzed from the perspective of function approximation. The influences of coregistration offset are also taken into consideration, which makes the conversion from the measurement baseline to the interferometric baseline complete.Thirdly, the parameter-saving spline representation model for the inter-satellite relative positioning based on the multi-frequency GNSS is established. On the one hand, the multi-frequency GNSS ambiguity resolution methods are raised tailored for the formation flying environment. The geometric relationship between the observation data and the estimated parameters is utilized, combining with the constraint of the continuous variation of the estimated parameters with epochs. Based on the function approximation and parameter-saving modeling theory, the relative positioning spline representation model is brought forward based on multi-frequency GNSS. The theoretical basis and simulation results of the relative positioning precision improvement are presented. The advantages of multi-frequency data over dual-frequency data and the ones of spline representation model over the traditional pointwise model are shown. On the other hand, the key issue of ambiguity resolution in the GNSS relative positioning is investigated. The performance diversity introduced by different conditions is illustrated, such as the geometry-free or geometry-based model, the original or combinational observations, and searching or sequential rounding strategies. A cascade tri-frequency ambiguity fixing method is presented, based on geometry-based model and rounding following integer decorrelation transformation. It is capable of instantaneous ambiguity fixing and robust to observation errors in the formation environment.Finally, the evaluation for the satellite spatial states determination performances based on multi-GNSS is given, and the navigation satellites selection criteria are raised oriented the distributed InSAR performance. The advantages of integrated GNSS are presented qualitatively. The quantitative performance evaluation indices are given for the satellite spatial states determination, such as the internal and external reliability, the success rate of carrier phase integer ambiguity resolution, and the observation geometry in particular. The comprehensive simulation examples and rigid theoretical foundations for the performance enhancement are presented. The redundancy and contributions of the observation data are considered synthetically. By combing the GNSS measurement sub-system and the InSAR sub-system, a novel selection criterion for the navigation satellites is raised from the aspect of elevation accuracy inversion, and higher elevation reconstruction accuracy is achieved under the same observation conditions.更多还原