模型误差与轨道精度

应用模型与实际模型总是存在差异,如何从理论上探讨这种差异的影响以及在实际处理中如何进行选择是非常困难的。本文中分析了差异的主要原因,明确了模型误差的概念,并从理论上探讨了模型误差对轨道估计精度的影响,给出了具体的影响误差计算公式;在此基础上,提出了轨道参数的估计效率和客户效率的概念,并由此导出了模型误差处理原则,从而回答了实际数据处理中,模型误差处理方案的选择问题。最后,给出了一个应用实例     

层次分析法中和积法的最优化理论基础及性质

和积法是一种计算排序向量的近似算法．本文从平方差均值平方根的概念出发,导出了和积法计算排序向量的显式表达式,为和积法作为一种排序方法提供了最优化方面的理论依据,并从保序性、相容性、对称性等角度对该方法的合理性进行了研究     

计算E型方杯压敏电阻最佳位置的一种有效方法

采用有限元法计算了Ｅ型方杯的挠度及应力分布,并从应力差计算出了压敏电阻设置的最佳位置．     

A MULTISTEP CHARACTERISTIC FINITE DIFFERENCE METHOD FOR TWO-DIMENSIONAL NONLINEAR CONVECTION-DIFFUSION PROBLEMS

1.IntroductionItiswellknowthattheconvection-dictionproblemsareakindofveryimportalprob-lems,particularlylthenonlinearconvection-diffusionproblexns.Inthepaper[1],asinglestepcharacteristicflutedmerencemethodisgivenforsolvingtwcadimensionalnonlinearconvection-dilhaionproblemswiththeconvectivecoefficientsdominating.HereweshallgiveamultistepcharactersticamrencemethodbasedonthelinearandquadraticinterpolationsforsolvingtwcadAnensionalnonlinearconvection-~ionproblemsinordertoincreasetheaccuracyofapprot…     

A SINGLESTEP CHARACTERISTIC FINITE DIFFERENCE METHOD FOR TWO-DIMENSIONAL NONLINEAR CONVECTION-DIFFUSION PROBLEMS

1.IntroductionThenonlinearconvection-diffusionproblemsareakindofveryimportantproblemssuchastheundergroundpercolation,air--pollution,fluidmechanicsmodelingetc.Thenumericalmethodsforsolvingthelinearproblemswiththeconvectivecoefficientsdominating,whicharethecharacteristicfiniteelementandcharacteristicfinitedifferencemethods,aregivenbyJ.DouglasandT.F.Russellin[1]andfurtherdevelopmentin[2]and[3],etc.ButtheconvergenceofmethodsneedstherestrictiveassumptionAt=O(h).Inpracticeonewouldliketotakebigger…     

合成孔径雷达景象匹配中制导导弹定位

研究了合成孔径雷达景象匹配中制导中的导弹定位问题。部分孔径成像是景象匹配中制导合成孔径雷达有效的成像模式,为便于利用多匹配点定位,提出了聚焦多普勒波束锐化成像方法。建立了基于匹配点距离差和多普勒差,并考虑了高度、速度和匹配点位置测量误差的定位模型。推导了位置估计的克拉美-罗下限（Cramer-Rao lower bound, CRLB）。提出了一种基于最小二乘的闭式求解算法。理论分析表明,在一定近似条件下,这一算法的估计精度到达CRLB。仿真分析了各种误差因素对CRLB的影响,验证了算法的理论分析结果及其有效性。     

基于四元数平方根UKF算法的SINS/GPS紧组合导航系统研究

在较大初始姿态误差角下,针对SINS/GPS紧组合导航系统扩展卡尔曼滤波（extenthed Kalman filter, EKF）算法定位精度下降的问题,提出了一种基于四元数的平方根无迹卡尔曼滤波（square root unscented Kalman filter, SRUKF）算法。为解决SRUKF算法中四元数正交规范化的限制,通过构造姿态矩阵代价函数将四元数预测均值问题转化为代价函数最小时的四元数向量求解问题,保证了均值四元数的规范化;利用乘性四元数误差表示四元数预测值与均值之间的距离,求取四元数的预测协方差矩阵,保证了算法的合理性。在此基础上,给出了SINS/GPS紧组合系统四元数平方根无迹卡尔曼滤波算法的具体步骤。在较大初始姿态误差角下的仿真实验结果表明,与EKF算法相比,该算法精度更高,稳定性更强。     

调制型光纤捷联系统系泊状态标校方法

针对舰船系泊状态特有的运动环境及惯性器件偏差对捷联系统导航精度的影响,提出一种单轴四位置旋转调制方案并建立该系统误差方程。研究了直观分析系统参数可观测性的解析方法,并运用谱条件数法计算出惯性测量单元（inertial measurement unit, IMU）在静止和旋转状态下各状态量的可观测度,利用分离位置回路的卡尔曼滤波模型实现了器件偏差的现场估计。数值仿真显示,单轴四位置转停方案可以有效地提高惯性器件偏差的可观测度,大大提高了系泊状态下惯性器件偏差的估计精度。     

曲面薄壳有限元离散误差分析和修正方法研究

基于曲面薄壳几何/物理方程和有限元离散原理推导,揭示了壳单元离散方法特点、刚度矩阵误差产生机理和范数表示方法。以航空涡轮发动机测压探针为例,深入研究了利用协调平板壳单元、非协调平板壳单元、一次和二次实体退化壳单元等不同类型单元离散曲面薄壳时,应力计算误差影响因素及变化规律,综述了单元的适用条件。基于Boltzmann模型提出了应力计算结果修正通式,通过FORTRAN语言编程修正应力仿真结果和ANSYS二次开发语言APDL命令实现应力修正结果的读取和后处理,应力修正结果与理论解相对误差小于5%,满足曲面薄壳设计和稳定性分析应用要求。

Abstract：

Based on the geometric physical equations and deduction of discrete process of finite element for curved thin-walled structure,characteristics of shell elements for discretion and discrete error for element stiffness matrix were analyzed theoretically,and the norm expression of discrete error for element stiffness matrix was put forward.Then,as an illustration,a probe for detecting the pressure inside turboprops was meshed by curved thin-walled shell element,compatible flat shell element,incompatible flat shell element and 3D degenerated shell element,respectively,and the stress of the probe was numerically calculated by FEM.On the basis of solved results,the stress error was analyzed thoroughly,the influence factors on the error and the applying situations for different elements were determined.Finally,the general form of correctional function was put forward according to the Boltzmann model.The program developed with FORTRAN was used to correct the numerical results and the APDL commands inside ANSYS software were utilized to read and plot the correctional stress results.Through the correction,the discrete error can be reduced within 5% that satisfies the requirement of application for structural design and stability analysis.     

基于IMU/GPS的性能导航方法研究

利用直角坐标系与大地坐标系之间的关系,将状态方程直接采用经纬度描述,GPS测量数据与电子地图上的经纬度相匹配,建立了混合模型,给出了一种新的GPS与IMU融合估计方法,提高了重复定位精度,更方便地实现组合导航系统功能。采用GPS和IMU装置进行了地面车辆试验和实际飞行试验,说明了模型和方法的适用性。

Abstract：

A mixed model was established and a new fusion estimation method for GPS and IMU was proposed through the relationship between the Cartesian coordinate system and the ground coordinate system,and the state equation described by the latitude and longitude directly,and the GPS measurement data and electronic map matched to the latitude and longitude.The provided method has improved the repeatability positioning accuracy and more easily integrated navigation system functions.The ground and actual flight were tested to illustrate the applicability of the model and method with GPS and IMU device.