关节铰间隙对漂浮基星载天线扰动研究

为了研究关节铰间隙对星载天线扰动,建立了铰间隙矢量矩模型,采用非线性弹簧阻尼及修正库仑摩擦建立了铰间隙碰撞力和摩擦力模型;进一步推导了理想铰和含间隙铰变拓扑结构的漂浮基星载天线多体系统动力学模型。通过接触碰撞判别准则,将系统的几何约束转化为力约束,使得系统无拓扑结构变化,便于系统的全局仿真。最后对理想铰与含间隙铰的漂浮基星载天线系统仿真分析,结果表明,较小间隙对星载天线位姿和指向影响很小,但极大增加关节的碰撞力,随着时间流逝,其位姿和指向精度偏差变大,但由于非线性阻尼的存在,碰撞力逐渐变小,最终保持接触状态。结论对星载天线指向精度的分析与控制具有重要的理论价值及工程实际意义。
     

卫星CMAC神经网络姿态控制器设计及仿真

为提高传统卫星姿态控制系统精度,提出了一种基于小脑模型(CMAC)神经网络的比例-积分-微分(PID)的复合控制器。给出了具在线学习功能的复合控制器结构,并证明了神经网络学习收敛条件与最终控制目标的一致性。仿真结果表明,设计的控制器具有较好的自适应性和鲁棒性。与传统控制器相比,进入稳定状态的速度更快,指向精度更高。     

基于交叠视场亮度优选算法的导航星库构建方法

星敏感器导航星库直接影响星图识别效率和姿态解算精度。分析了参与姿态解算的星颗数和亮度对姿态精度影响,在此基础上,提出了一种基于交叠视场亮度优选算法的星敏感器导航星库构建方法。将精度较高的Hipparcos星表作为基础星表,分析了星等、双星、自行、变星等对姿态精度的影响,将星库筛选成一个备选星库,生成覆盖全天球的交叠视场。每个拥有较多星的视场,都按照亮度优先原则从视场内的扇形区域中选择导航星,从而得到分布均匀的导航星库。结果表明:该方法能有效减小星敏感器导航星库规模,实现导航星在全天球和局部天区的均匀分布。     

In-orbit assessment of laser retro-reflector efficiency onboard high orbiting satellites

The navigation and geodetic satellites that orbit the Earth at altitudes of approximately 20,000 km are tracked routinely by many of the Satellite Laser Ranging (SLR) stations of the International Laser Ranging Service (ILRS). In order to meet increasing demands on SLR stations for daytime and nighttime observations, any new mission needs to ensure a strong return signal so that the target is easily acquirable. The ILRS has therefore set a minimum effective cross-section of 100 million square metres for the on-board laser retro-reflector arrays (LRAs) and further recommends the use of ‘uncoated’ cubes in the arrays. Given the large number of GNSS satellites that are currently supported by SLR, it is informative to make an assessment of the relative efficiencies of the various LRAs employed. This paper uses the laser ranging observations themselves to deduce and then compare the efficiencies of the LRAs on the COMPASS-M1 navigation satellite, two satellites from the GPS and three from the GLONASS constellations, the two GIOVE test satellites from the upcoming Galileo constellation, the two Etalon geodetic spheres and the geosynchronous communications test satellite, ETS-8. All the LRAs on this set of satellites employ back-coated retro-reflector cubes, except those on the COMPASS-M1 and ETS-8 vehicles which are uncoated. A measure of return signal strength, and thus of LRA-efficiency, is calculated using the laser-range full-rate data archive from 2007 to 2010, scaled to remove the effects of variations in satellite range, atmospheric attenuation and retro-reflector target total surface area. Observations from five SLR stations are used in this study; they are Herstmonceux (UK), Yarragadee (Australia), Monument Peak and McDonald (USA) and Wettzell (Germany). Careful consideration is given to the treatment of the observations from each station in order to take account of local working practices and system upgrades. The results show that the uncoated retro-reflector cubes offer significant improvements in efficiency.     

Determination of station positions and velocities from laser ranging observations to Ajisai,Starlette and Stella satellites

The positions and velocities of the four Satellite Laser Ranging (SLR) stations: Yarragadee (7090), Greenbelt (7105), Graz (7839) and Herstmonceux (7840) from 5-year (2001–2005) SLR data of low orbiting satellites (LEO): Ajisai, Starlette and Stella were determined. The orbits of these satellites were computed from the data provided by 20 SLR stations. All orbital computations were performed by means of NASA Goddard’s GEODYN-II program. The geocentric coordinates were transformed to the topocentric North–South, East–West and Vertical components in reference to ITRF2005. The influence of the number of normal points per orbital arc and the empirical acceleration coefficients on the quality of station coordinates was studied. To get standard deviation of the coordinates determination lower than 1 cm, the number of the normal points per site had to be greater than 50. The computed positions and velocities were compared to those derived from LAGEOS-1/LAGEOS-2 data. Three parameters were used for this comparison: station coordinates stability, differences from ITRF2005 positions and velocities. The stability of coordinates of LEO satellites is significantly worse (17.8 mm) than those of LAGEOS (7.6 mm), the better results are for Ajisai (15.4 mm) than for Starlette/Stella (20.4 mm). The difference in positions between the computed values and ITRF2005 were little bit worse for Starlette/Stella (6.6 mm) than for LAGEOS (4.6 mm), the results for Ajisai were five times worse (29.7 mm) probably due to center of mass correction of this satellite. The station velocities with some exceptions were on the same level (≈1 mm/year) for all satellites. The results presented in this work show that results from Starlette/Stella are better than those from Ajisai for station coordinates determination. We can applied the data from LEO satellites, especially Starlette and Stella for determination of the SLR station coordinates but with two times lower accuracy than when using LAGEOS data.     

弹道跟踪数据野值剔除方法性能分析

弹道跟踪数据中的野值会影响导弹目标定位精度。对用于弹道跟踪数据的野值剔除方法进行了综述。结合事后和实时两大类处理模式,分析了各野值剔除方法在理论和工程应用上的优缺点,给出了实际应用时的限制因素,建立了野值判决准则和相应的野值剔除实现步骤。仿真结果表明:野值剔除方法可以有效提高弹道跟踪数据的精度。事后处理模式中,基于格拉布斯准则的野值剔除效果最佳;实时处理模式中,基于自适应门限的五点线性预报法野值剔除效果相对较好,且时间复杂度相对较低,适合实时处理。     

脉冲激光测距误差标定及不确定度分析CSCD

以脉冲激光测距机的测距误差标定为背景,通过对测距误差解析表达式的推导,给出了系统各不确定度分量对测距误差的影响,计算了信号延迟时间、探测器和激光二极管响应时间、晶振频率、大气折射率等不确定度分量。通过对某已标定激光测距机实验验证表明:该装置在500m^20 000m测距范围内,测量重复性引入的相对标准不确定度最大值为0.67m,满足该类激光测距机测量不确定度5m(k=2)的校准测试需求。因此,利用该解析表达式可以实现对脉冲激光测距误差的有效评估,这对于脉冲激光测距测试系统、脉冲激光测距机的设计具有重要的指导意义。     

Determination of optimal samples for robot calibration based on error similarity

Industrial robots are used for automatic drilling and riveting.The absolute position accuracy of an industrial robot is one of the key performance indexes in aircraft assembly,and can be improved through error compensation to meet aircraft assembly requirements.The achievable accuracy and the difficulty of accuracy compensation implementation are closely related to the choice of sampling points.Therefore,based on the error similarity error compensation method,a method for choosing sampling points on a uniform grid is proposed.A simulation is conducted to analyze the influence of the sample point locations on error compensation.In addition,the grid steps of the sampling points are optimized using a statistical analysis method.The method is used to generate grids and optimize the grid steps of a Kuka KR-210 robot.The experimental results show that the method for planning sampling data can be used to effectively optimize the sampling grid.After error compensation,the position accuracy of the robot meets the position accuracy requirements.     

Accuracy enhancement of the spherical actuator with a two-level geometric calibration method

This paper presents a two-level geometric calibration method for the permanent magnet （PM） spherical actuator to improve its motion control accuracy. The proposed actuator is com- posed of a stator with circumferential coils and a rotor with multiple PM poles. Due to the assembly and fabrication errors, the real geometric parameters of the actuator will deviate from their design values. Hence, the identification of such errors is critical for the motion control tasks. A two-level geometric calibration approach is proposed to identify such errors. In the first level, the calibration model is formulated based on the differential form of the kinematic equation, which is to identify the geometric errors in the spherical joint. In the second level, the calibration model is formulated based on the differential form of torque formula, which is to calibrate the geometric parameters of the magnetization axes of PM poles and coils axes. To demonstrate the robustness and availability of the calibration algorithm, simulations are conducted. The results have shown that the proposed two-level calibration method can effectively compensate the geometric parameter errors and improve the positioning accuracy of the spherical actuator.     

中国深空网首次△DOR联合测轨试验分析

通过分析中国深空网首次△DOR（Delta Differential One way Ranging,双差分单向测距）联合测轨试验的干涉测量事后数据,重点从观测量随机精度、闭合时延等方面讨论了国内深空网与国内VLBI（Very Long Baseline Interferometry,甚长基线干涉测量）观测网、国内深空网与国际深空网的联合干涉处理情况,并与ESOC（European Space Operation Center,欧洲空间操作中心）数据处理结果进行了比对.试验结果表明：我国深空网已具备独立或联合开展深空探测器导航测轨的系统支持能力;深空站系统具备高速率数据接收、采集、记录、传输能力,采集数据处理精度优于1 ns;深空网干涉测量信号处理中心具备多体制信号的干涉处理分析能力,其分析精度与ESOC处理精度差异在0.1 ns量级.