卫星集群概念研究

卫星星座的核心特征是扩展覆盖,卫星编队飞行的核心特征是扩展基线,这些相关概念都已找到其突出优势的应用。模块化分离卫星的重点落在了卫星上,容易引起＂为什么用一群卫星取代一颗卫星＂的质疑,对认识集群技术、寻找其有突出优势的应用有误导作用。文章提出了卫星集群的概念,指出集群的组成是个性化的卫星而不是卫星分系统这一事实,重点应落在集群上;提出了卫星集群的核心特征是卫星之间的集群关系,包括信息、物质和能量等的自组织交换;提出了集群关系、个性化特征和进化能力是卫星集群的三个社会化基本特征;指出区分一颗卫星是否属于卫星集群的标准是看该卫星是否具有社会化特征;比对人类社会个体间相互联系,预测和分析了卫星集群的社会层次关系;指出卫星集群族级应用是突破口,为寻找卫星集群相对于单颗卫星具有突出优势的典型应用指出了思路和方向。     

Characteristics and accuracies of the GRACE inter-satellite pointing

For almost 10 years, the Gravity Recovery and Climate Experiment (GRACE) has provided information about the Earth gravity field with unprecedented accuracy. Efforts are ongoing to approach the GRACE baseline accuracy as there still remains an order of magnitude between the present error level of the gravity field solutions and the GRACE baseline. At the current level of accuracy, thorough investigation of sensor related effects is necessary as they are one of the potential contributors to the error budget. In the science mode operations, the twin satellites are kept precisely pointed with their KBR antennas towards each other. It is the task of the onboard attitude and orbit control system (AOCS) to keep the satellites in the required formation. We analyzed long time series of the inter-satellite pointing variations as they reflect the AOCS performance and characteristics. We present significant systematic effects in the inter-satellite pointing and discuss their possible sources. Prominent features are especially related to the magnetic torquer characteristics, star cameras’ performance and KBR antenna calibration parameters. The relation between the magnetic torquer attitude control and the Earth magnetic field, impact of the different performance of the two star camera heads on the attitude control and the features due to uncertainties in the calibration parameters relating the star camera frame to K-frame are discussed in detail. Proper understanding of these effects will help to reduce their impact on the science data and subsequently increase the accuracy of the gravity field solutions. Moreover, understanding the complexity of the onboard system is essential not only for increasing the accuracy of the GRACE data but also for the development of the future gravity field satellite missions.     

A novel system parameters design and performance analysis method for Distributed Satellite-borne SAR system

As a promising new technology emerged in recent years, the Distributed Satellite-borne SAR (DSS) system with Interferometric Synthetic Aperture Radar (InSAR) imaging capability has been recognized by the remote sensing community as an integrated part of the spaceborne earth observation system. However, most researches on DSS were focused on individual aspects of system design/analysis and data processing, while few studies have been dedicated to establishing a standard methodological framework for universal DSS system design. Aiming at this problem, the topics of DSS system error analysis and design method are investigated in this paper. Firstly, a rigorous error propagation model of height measurement is theoretically derived from the DSS InSAR imaging geometry, and the impact of each error sources in this model on the height measurement accuracy is analyzed individually. In particular, the baseline length and its measurement error in the InSAR imaging plane are identified as the dominant factors. Second, a new method for DSS system design is proposed based on our analyses with the error propagation model. This method consists of two important phases, namely collaboration design and monostatic design. The major objective of collaboration design is to determine the configuration and parameters for the coordination between satellites in the DSS system. Afterwards, the key parameters for single satellite and the payload are determined by the monostatic design. Thirdly, a system performance analysis method is developed to comprehensively evaluate the performance of the pre-designed DSS system. In the next section, a typical example of DSS system design is given to demonstrate the effectiveness of the proposed system design and analysis methods. The target DSS system has the same goal and hardware configuration as the TanDEM-X mission. Finally, the conclusion is drawn with our major findings are presented.     

采用程序集技术的卫星数据解析软件设计

卫星实时定位和事后数据精度分析中,不同厂家的卫星接收机经常一起使用,同一种类数据帧格式彼此不同,甚至同一厂家、不同型号的卫星接收机数据格式也经常变化,这给数据解析带来很大不便。本文在卫星接收机数据解析软件中采用面向对象的设计方法,利用UML（统一建模语言）进行建模,针对数据格式经常变化的情况采用数据库预定义技术,最后通过.Net程序集技术实现了不改变软件代码即可适应新的卫星接收机数据格式的目标。实际应用证明,该软件具有结构直观、处理数据种类多、可扩展性强等特点,在卫星数据解析的通用性方面做了有益的尝试。     

Image Quality assessment for Chang＇E-1 IIM 2C Data

嫦娥一号干涉成像光谱仪（IIM）获取的多波段图像数据对于月表矿物成分反演具有重要意义.从目视质量、客观参数、几何配准精度、图像质量随时间的变化情况等方面对IIM 2C级的多波段图像数据质量进行了评价.结果显示,其两端波段范围的图像质量较差,不能有效识别月表单元,中间波段图像质量较好;自然色图像显示IIM数据存在明显的横向色调不均一性.利用图像的客观评价指标（辐射精度、清晰度、信息量）将干涉成像光谱仪图像数据与美国Clementine UV-VIS多光谱图像数据进行对比分析,结果表明,IIM数据清晰度好于ClementineUV-VIS,而在方差和信息量上弱于Clementine UV-VIS.将IIM图像数据与Clementine多光谱数据统一到相同的空间分辨率,进行配准精度评价,其最大控制点配准误差为0.63 pixel.此外,还对时间跨度两个月覆盖月表同一区域的IIM图像数据从辐射精度、信息量和清晰度方面进行了评价.     

应用于卫星自主任务调度的改进遗传算法

针对具有侧摆能力的对地观测卫星的自主任务调度问题,对卫星自主任务调度问题和约束条件进行了描述,针对卫星自主任务调度NP-hard的特点,构建了基于目标收益及多约束卫星任务调度模型。设计了一种改进的遗传算法,从遗传操作的各个部分进行算法优化。首先将小区间法应用于初始种群生成,保证了种群的多样性,并且交叉和变异算子均引入自适应概率;同时采用两代竞争技术来避免“早熟”现象,提高算法的效率和鲁棒性。算法还采用最优保留策略用来保存进化中的最优解,使得算法收敛于全局最优。对局部多冲突观测任务应用该改进遗传算法,并针对区域密集目标的观测问题设计了仿真试验,与传统模拟退火算法及免疫蚁群遗传混合算法进行了比较,验证了该算法的有效性和收敛效果。     

动态拖尾星图模拟算法研究

星图模拟技术是星图识别算法仿真和性能测试的基础.为了更好地模拟星敏感器在轨工作的情况,对星敏感器的软件开发以及自身性能进行有效的地面测试,提出一种动态拖尾星图模拟算法.该算法分为4个步骤:首先根据星敏感器指向在全天球范围内搜索导航星;接着利用小孔成像模型计算导航星在星敏感器成像面的投影位置;然后在考虑卫星运动引起的恒星拖尾的基础上,按照二维灰度分布规律置灰度值来模拟星像点像素;最后叠加杂散光背景及成像器件引起的噪声.提出的算法具有速度快、精度高及可实现性高等优点.     

无阻力双星编队的满系数矩阵MIMO定量反馈控制

研究由两颗无阻力卫星构成的、用于重力场测量的松散式编队相对位置控制方法,主要解决编队控制所产生的非重力加速度在重力场测量频带内的干扰抑制问题。首先,选取双星位置中点作为编队系统质心,建立了考虑J2摄动项的双星相对动力学模型。然后,根据定量反馈理论（QFT）确定系统在频域内的跟踪性能、鲁棒稳定性、输入干扰抑制等约束。与当前常规的对角型QFT控制器设计方法不同,本文针对编队系统的多输入多输出（MIMO）通道强耦合特性,设计了更具一般性的满系数矩阵鲁棒控制器,不但实现了闭环控制回路整定、通道解耦和稳态收敛,还有效抑制了编队控制量功率谱在科学测量频带内的干扰。最后,通过在时域中的数字仿真校验了该方法控制器的有效性和鲁棒性。     

柔性卫星角加速度轨迹规划与振动抑制方法

针对柔性卫星在快速机动过程中会激发柔性附件振动的问题,提出一种通过设计角加速度轨迹来抑制振动的方法。在存在振动频率不确定性的情况下,所规划的角加速度轨迹能够将残余振动幅值抑制在指定的范围内。类比不同幂次余弦函数激发的残余振动表达式,设计一种含有多个待定系数的角加速度曲线,并根据残余振动幅值对待定系数进行约束,从而求解出所有的待定系数。这种设计方法改进了余弦函数抑制低频振动需要增加较大机动时间的问题。相比输入成型方法,所规划的轨迹光滑,增加的机动时间较短,对高频振动有着更好的抑制效果。     

尖楔前体飞行器FADS系统驻点压力对神经网络算法精度的影响

针对尖楔前体（类乘波体）飞行器用嵌入式大气数据传感（FADS）系统存在建模困难及解算模型精度低的问题,首先采用BP神经网络建模代替传统的FADS系统空气动力学建模的方法,建立含有双隐含层的四层神经网络模型,然后通过合理选择网络结构参数及训练验证,对FADS系统的攻角进行解算,最后对驻点压力对算法精度的影响进行研究。结果表明,本文建立的含有双隐含层的四层神经网络模型精度较高,攻角测试误差小于 0.25°; 驻点压力是否作为输入参数对FADS系统神经网络算法求解精度影响较大,攻角测试误差相差达0.1°。在飞行器前缘半径允许的情况下,应尽量得到驻点压力用于解算攻角,提高解算精度。