基于小波变换与滤波的DEM简化方法

针对现有3维地形模型简化方法中的不足,提出了一种基于小波变换与滤波的规则格网简化方法。同时还在大量实验的基础上,给出了一般情况下需要的变换层数,分析了滤波门限与简化效果之间的关系。这里的研究不仅为3维地形的多分辨率合理建模与精度评估提供了理论和方法,而且也为今后开展类似研究及进行建模质量控制提供了科学的依据。     

Chebyshev逼近滤波器在位场分离中的应用

在对经典FIR数字滤波器的设计方法进行研究的基础上,提出了一种可以用于位场分离的基于Chebyshev最佳一致逼近原理的FIR滤波器的设计方法。在理论模型实验中,采用基于Hanning窗的低通滤波器计算出的区域异常最大误差为6.266×10^-6 m/s² ,均方差为2.115×10^-6 m/s² ,最大百分比误差为22.2%,而且计算点在±9 km以外的误差均大于10.1%。而利用最佳一致逼近原理分离出的区域场和局部场与理论异常值拟合得较好,曲线基本重合。分离出的区域异常最大误差为3.101×10^-6 m/s² ,均方差为0.989×10^-6 m/s² ,最大百分比误差仅在边部的几个数据上,为7.76%,其余各点的误差均小于4.1%。实例检验中将该方法用于孙吴—嘉荫剖面布格重力异常场的分离,分离出的区域场中局部场残留少,分离彻底,效果较为理想。     

一种新型非线性滤波理论及其应用

研究一种新型的非线性滤波理论,即Unscented卡尔曼滤波（UKF）,同时为了获得更高的计算效率和确保协方差阵的非负定性,研究了平方根UKF。将UKF和平方根UKF应用到星载GPS卫星定轨中,实际算例表明UKF和平方根UKF的性能要优于常用的推广卡尔曼滤波的性能。     

Bringing GPS into harsh environments for fully automated deformation monitoring

Engineering projects that require deformation monitoring frequently utilize geodetic sensors to measure displacements of target points located in the deformation zone. In situations where control stations and targets are separated by a kilometer or more, GPS can offer higher precision position updates at more frequent intervals than can normally be achieved using total station technology. For large-scale deformation projects requiring the highest precision, it is therefore advisable to use a combination of the two sensors. In response to the need for high precision, continuous GPS position updates in harsh deformation monitoring environments, a software has been developed that employs triple-differenced carrier-phase measurements in a delayed-state Kalman filter. Two data sets were analyzed to test the capabilities of the software. In the first test, a GPS antenna was displaced using a translation stage to mimic slow deformation. In the second test, data collected at a large open pit mine were processed. It was shown that the delayed-state Kalman filter developed could detect millimeter-level displacements of a GPS antenna. The actual precision attained depends upon the amount of process noise infused at each epoch to accommodate the antenna displacements. Higher process noise values result in quicker detection times, but at the same time increase the noise in the solutions. A slow, 25 mm displacement was detected within 30 min of the full displacement with sigma values in E, N and U of ±10 mm or better. The same displacement could also be detected in less than 5 h with sigma values in E, N and U of ±5 mm or better. The software works best for detecting long period deformations (e.g., 20 mm per day or less) for which sigma values of 1–2 mm are attained in all three solution components. It was also shown that the triple-differenced carrier-phase observation can be used to significantly reduce the effects of residual tropospheric delay that would normally plague double-differenced observations in harsh GPS environments.

An approach to aid INS using time-differenced GPS carrier phase (TDCP) measurements

This paper presents a simple and effective approach that incorporates single-frequency, L1 time-differenced GPS carrier phase (TDCP) measurements without the need of ambiguity resolution techniques and the complexity to accommodate the delayed-state terms. Static trial results are included to illustrate the stochastic characteristics and effectiveness of the TDCP measurements in controlling position error growth. The formulation of the TDCP observation model is also described in a 17-state tightly-coupled GPS/INS iterative, extended Kalman filter (IEKF) approach. Preliminary land vehicle trial results are also presented to illustrate the effectiveness of the TDCP which provides sub-meter positional accuracies when operating for more than 10 min.     

经验模式分解的尺度阈值滤波在变形分析中的应用

本文将信号处理领域的经验模式分解算法应用于变形信息的提取中,通过引入阈值函数,建立了基于经验模式分解的尺度阈值滤波模型,采用优化模型确定了经验模式分解的次数。分别通过模拟试验和实测数据与小波阈值法和多项式拟合法进行了比对,分析表明:在低噪声情况下,三种方法都有一定的滤波效果;在高噪声情况下,经验模式分解的尺度阈值滤波法具有与小波阈值去噪法相等的精度,而且瞬时强噪声识别能力更好,优于多项式拟合法。     

有色噪声作用下的卡尔曼滤波

利用GPS载波相位三差观测量进行动态定位(或精密导航),就必须研究有色噪声滤波的理论问题.根据需求,推导了动态噪声、观测噪声为有色噪声的线性系统滤波公式,并证明白噪声卡尔曼滤波是有色噪声卡尔曼滤波的特例,或者说有色噪声的卡尔曼滤波是白噪声卡尔曼滤波的推广.     

UKF算法及其在GPS卫星轨道短期预报中的应用

分析了GPS卫星预报星历,在比较分析EKF和UKF优缺点的基础上,将UKF引入GPS卫星轨道预报研究中.数值模拟和结果分析表明,UKF方法预报更稳定,能有效地提高轨道预报精度和稳定性.     

基于增强型GPS的自适应UKF实时星间相对定位方法

高精度的星间实时相对定位是卫星编队飞行的一项关键技术。本文以双星编队为例,提出一种基于GPS双频P码、双频载波相位以及星间距离观测信息的自适应UKF、实时相对定位方法。仿真结果表明:相比传统的EKF方法,该算法能有效地提高滤波稳定性及相对定位精度,而且,星间测距信息的引入能大大减少整周模糊度判定所需的历元数。     

基于ICCP算法的重力辅助惯性导航

迭代最近等值线算法(ICCP)是一种重要的匹配导航算法,文中首先介绍ICCP算法的基本原理,随后在0.2′×0.2′重力异常数据库的基础上,利用ICCP算法进行仿真计算得到最佳匹配位置。最后为了验证匹配位置是否可用于修正惯导误差,提出将匹配位置误差作为观测量,用卡尔曼滤波对惯导系统误差进行最优估计。由最后的仿真结果可以看出,ICCP算法可有效抑制惯导纬度误差的增长,且最大纬度误差不超过2,′以匹配位置误差作为观测量可以用来估计惯导方位误差角。