基于Allan方差的陀螺随机误差分析

随机误差是陀螺误差分析中重要的一部分,将Allan方差用于陀螺随机误差分析中,首先介绍了Allan方差的计算步骤,其次详细论述了随机误差的类型及其在Allan方差双对数图中的表现形式,最后通过仿真以及实测数据进行验证,结果表明Allan方差可以对各种误差源的统计特性进行辨识,是一种简单实用的陀螺随机误差识别方法。     

基于RWTLS的Allan方差算法简化

Allan方差是现在应用最广泛的随机误差辨识方法之一。大量的试验表明Allan方差可有效地分离出导航过程中的多项随机误差,但是Allan方差也有自身的局限性。针对Allan方差在处理大数据量时计算效率低下、辨识度受粗差的影响较大的问题,本文提出了简化Allan方差算法的方案。首先,在确保Allan方差计算准确的前提下,以提高Allan方差计算效率为目的,对Allan方差算法进行简化;然后,利用抗差加权整体最小二乘（RWTLS）模型的迭代算法对简化后的Allan方差辨识结果进行抗差拟合处理;最后,以光纤式惯性测量单元（IMU）为分析对象,设计试验方案对简化后的Allan方差进行验证。     

基于Allan方差的BDS随机误差分析

BDS观测数据噪声的辨识对提高BDS定位精度具有非常重要的意义,大量研究表明GPS非差随机模型并不服从高斯白噪声分布,北斗系统由于其星座和频率的多样性,噪声特征将明显不同于GPS。本文利用iGMAS大量高频实测数据,通过Allan方差分析,对比研究了BDS的噪声成分与其参数。试验结果表明,BDS非差噪声成分与GPS相同,主要为白噪声与一阶马尔科夫过程噪声,但是具体参数不同。BDS相位白噪声为2.50 mm,大于GPS的2.33 mm;GM过程噪声为2.38 mm·s^-1/2,小于GPS的4.43 mm·s^-1/2,相关时间为77.07 s,大于GPS的55.51 s。大量试验数据表明BDS的噪声参数聚集程度大于GPS,对环境的敏感程度低,噪声相关时间更长,过程噪声更小,这在北斗精密定位中还有待精化。     

基于Allan方差和SVR的MEMS陀螺仪随机误差分析与预测

为了使低成本MEMS陀螺仪数据的精度更高,本文提出了一种混合核函数支持向量回归（SVR）的MEMS陀螺仪随机误差预测模型,并通过粒子群优化（PSO）算法对模型参数和核函数参数进行优化;同时通过Allan方差法对SVR预测前后的MEMS陀螺仪随机误差数据进行分析。试验结果表明：混合核函数SVR对MEMS陀螺仪随机误差的预测准确度可达99.99%;当MEMS陀螺仪所处状态不同,但噪声特性相同时,可采用统一的SVR预测模型预测随机误差,研究结果为进一步用于MEMS陀螺仪的实时误差补偿中提供依据。     

运动状态下MEMS IMU随机误差的Allan方差分析方法

在车载全球卫星导航系统（global navigation satellite system,GNSS）/微机械系统（micro-electro mechanical system,MEMS）惯性测量单元（inertial measurement unit,IMU）组合导航滤波解算时,通常使用MEMS厂商标定的加速度计和陀螺仪的随机模型参数（简称为标称参数）。这些标称参数由传感器厂商在静止状态或通过实验室转台设备来测定,是否适用于运动状态下的组合导航滤波解算并获得最优的定位精度,需要进一步研究和评估。本文提出了一种运动状态下MEMS IMU随机误差的Allan方差分析方法,将不同精度等级的IMU安装在同一车载平台上同步采集观测数据,用高精度IMU观测数据提取车辆运动信息,然后从低精度MEMS IMU观测数据中剔除车辆运动信息得到类似静止的观测数据,进行Allan方差分析,获得运动状态下的MEMS随机模型参数,并将其应用于GNSS/惯性导航系统（inertial navigation system,INS）组合导航解算。试验结果表明,采用运动状态下标定的随机模型参数,组合导航的定位精度优...     

基于OpenMP的海量DEM快速并行裁切算法

为了提高海量DEM的裁切效率,且充分利用现有的多核CPU计算资源,提出一种基于OpenMP框架的快速并行裁切算法,并对其性能进行实例验证.通过对规则和不规则DEM数据用不规则矢量范围线裁切的任务进行分解,设计了相应的并行程序,将遍历点与矢量范围线的空间关系及DEM内插的过程进行并行化计算,缩短了计算时间.算法在实际生产...     

基于集群的影像并行OPTA算法研究

并行计算是提高海量遥感影像细化处理速度的有效途径.基于对OPTA细化算法的分析,深入研究OPTA细化并行算法中数据划分和数据通信优化等关键问题,提出一种适合集群并行处理的通信优化并行算法,实验证明该算法具有良好的加速比,适合集群计算环境下海量遥感影像的细化处理.     

集群环境下的影像并行匹配算法

遥感影像的快速匹配是3维信息实时获取的关键,而采用并行处理技术能够有效地解决匹配的速度问题.在对基于概率松弛的匹配算法进行分析的基础上,结合集群并行计算平台的特点,从负载平衡和数据通信两个方面出发,对并行匹配算法中的数据划分关键问题进行了深入讨论,提出一种负载平衡的并行匹配算法.实验结果证明,该算法能够达到计算节点的负载平衡,通信开销更小,适合在集群计算平台进行高性能影像匹配处理.     

自然资源分区矢量数据的快速并行统计方法

针对自然资源数据统计计算所面临的高耗时问题,该文提出了一种支持大规模分区矢量数据的快速并行统计方法。基于该方法构建的高性能计算系统可以解决分布式数据组织、集群计算和统计模型管理3方面问题,从而为利用高性能计算设施提升统计计算效率提供一种有效手段。实验以表面面积和地表覆盖转移矩阵为例,在集群环境下完成全国陆域全要素数据计算测试,性能提升分别达到113.95倍与65.56倍,综合验证了该文方法的性能优势。     

Using Allan variance to analyze the error characteristics of GNSS positioning

Currently, we evaluate the positioning accuracy of GNSS mainly by providing statistical values that can represent the overall error level, such as CEP, RMS, 2DRMS, and maximum error. These are solid indicators of the general performance of the GNSS positioning. But some applications like GNSS/INS integrated system require a detailed analysis of the error characteristics and knowledge of the precise error model. This requirement necessitates the modeling of the error components of the GNSS positioning solutions. In our research, the Allan variance method is proposed to analyze the GNSS positioning errors, describe the error characteristics, and build the corresponding error models. Based on our research, four dominant noise terms are identified in the GNSS positioning solutions, that is, 1st order Gauss-Markov process, Gaussian white noise, random walk noise, and flicker noise, which indicates that white noise is not always enough and appropriate to model GNSS positioning errors for some applications. The results show that the Allan variance is a feasible and effective way to analyze the error characteristics of the GNSS positioning solutions.     

Using Allan variance to improve stochastic modeling for accurate GNSS/INS integrated navigation

MEMS-based integrated system of a global navigation satellite system (GNSS) and an inertial navigation system (INS) has been widely used in various navigation applications. However, such integration encounters some major limitations. On the one hand, the noisy MEMS-based INS undermines the accuracy with time during the frequently occurring GNSS outages caused by signal blockage or attenuation in certain situations such as urban canyon, tunnels, and high trees. On the other hand, the model mismatch between actual GNSS error and the assumed one would also degrade the obtained accuracy even with continuous GNSS aiding. To improve the overall performance for GNSS/MEMS-INS, better error models can be obtained using Allan variance (AV) analysis technique for modeling inertial sensor errors instead of the commonly recommended auto-regressive processes, and on the other hand, the measurement update in Kalman filter is improved using innovation filtering and AV calculation. The performance of each method and the combined algorithm is evaluated by a field test with either differential GNSS (DGNSS) or single-point positioning (SPP) as external aid. In addition to the considerable navigation enhancement brought by each method, the experimental results show the combined algorithm accomplishes overall accuracy improvements by about 18% (position), 8% (velocity), and 38% (attitude) for integration with DGNSS, and by about 15% (position), 75% (velocity), and 77% (attitude) for that with SPP, compared with corresponding traditional counterparts.     

快速生成等值线的凹包算法

针对传统等值线的生成是根据离散点利用凸包作为构建TIN的边界,采用等值线追踪的方法进行生成,然后再进行等值线光滑、添加相应注记等以完成等值线图的绘制,存在效率不高,待删除的冗余三角形较多等问题,该文提出了一种以凹包作为由离散点构建TIN的边界,将所有三角形中等值线段生成之后,再整体进行相邻线段合并成多段线折线的方法进行等值线生成,减少了冗余三角形。经过实际验证表明,此法不仅能够满足等值线的快速生成,而且对于狭长型地带或大数据量时其优势更加明显,能够有效降低劳动强度,并且此法简单易行,能够满足实际的生产需要。     

结合边缘编组的Hough变换直线提取

针对传统Hough变换用于直线检测存在的问题进行了细致的分析和归纳总结,在此基础上,提出一种结合边缘编组的Hough变换直线提取算法。该算法首先采用基于8邻域的边缘跟踪算法对Canny算子检测得到的边缘点进行编组;然后对每一个边缘组分别进行Hough变换,单独确定Hough变换原点和参数的取值范围。Hough变换过程中,采用迭代的"投票"方式,每次确定单一峰值点并删除对应像素。实验证明,该算法原理简单,能有效解决传统Hough变换存在的精度不高、计算复杂等问题。同时该算法具有较强的鲁棒性,可以有效处理不同类型的影像数据,适用于并行处理。     

基于GPGPU的全波形并行分解算法

针对EM(Expectation Maximization)波形分解算法具有多次迭代和大量乘、除、累加等高密集运算的特点,提出一套将EM算法在通用计算图形处理器GPGPU上并行化的方案。针对通用并行计算架构CUDA的存储层次特点,设计总体的并行方案,充分挖掘共享存储器、纹理存储器的高速访存的潜能;根据波形采样值采用字节存储的特征,利用波形采样值的直方图求取中位数,从而降低求噪音阈值的计算复杂度;最后,采用求和规约的并行策略提高EM算法迭代过程中大量累加的计算效率。实验结果表明,当设置合理的并行参数、EM迭代次数大于16次、数据量大于64 M时,与单核CPU处理相比,GPU的加速比达到了8,能够显著地提高全波形分解的效率。     

可扩展的迭代自组织分析并行处理算法

随着遥感影像数据量的增加,传统非监督分类迭代自组织分析(ISODATA)算法的运算将十分耗时,应用并行计算技术能够有效解决该性能瓶颈。针对现有基于并行计算模型MapReduce的遥感迭代自组织分析并行算法存在的局限性,提出一种可扩展的基于MapReduce的迭代自组织分析并行处理算法。该算法通过其包含的全局子采样算法、聚类中心点集合过滤算法以及聚类映射算法,有效克服了现有并行算法中存在的不足。实验结果表明,在同等规模遥感计算中,该算法效率高于现有并行处理算法,具有良好的加速比,且在处理更大的影像块时具有更高的精度。     

Recursive algorithm for fast GNSS orbit fitting

Gaussian elimination is an efficient and numerically stable algorithm for estimating parameters and their precision. However, before estimating the parameters, it is often prudent to perform statistical tests to achieve the best fitting model. We use Gaussian elimination to select the best fitting model among candidate models. A succinct relationship between the weighted sum of squared residuals and the previous one is revealed by a volume formula. For quick parameter estimation and determination of weighted sum of squared residuals, a recursive elimination algorithm is proposed in the context of Gaussian elimination. In order to improve the model selection efficiency, the parameter estimation and the determination of the weighted sum of squared residuals are carried out in parallel using the proposed recursive elimination algorithm in which the improvement at each recursive stage is judged by the Bayesian information criterion. Ultimately, the computational complexity and numerical stability of the recursive elimination proposed are briefly discussed, and a GNSS orbit interpolation example is used to verify the results. It shows that the proposed recursive elimination algorithm inherits the numerical stability of the Gaussian elimination, and this algorithm can be used to examine the gain from the newly introduced parameter, dynamically assess the fitting model, and fix the optimal model efficiently. The optimal fitting model with the lowest information is very close to the real situation verified by checkpoints.     

A fast algorithm of cartographic sounding selection

An effective strategy and framework that adequately integrate the automated and manual processes for fast cartographic sounding selection is presented. The important submarine topographic features are extracted for important soundings selection, and an improved “ influence circle” algorithm is introduced for sounding selection. For automatic configuration of soundings distribution pattern, a special algorithm considering multi-factors is employed. A semi-automatic method for solving the ambiguous conflicts is described. On the basis of the algorithms and strategies a system named HGIS for fast cartographic sounding selection is developed and applied in Chinese Marine Safety Administration Bureau （CMSAB）. The application experiments show that the system is effective and reliable. At last some conclusions and the future work are given.     

基于CUDA的并行空间计算

基于GPGPU的并行计算是目前新兴的热门研究领域。针对空间信息处理中常涉及的算法,本文讨论了基于CUDA技术的算法并行化问题,并提出基于栅格的点集凸壳并行算法及并行耳三角剖分算法。经对实际数据的实验表明,基于CUDA的几何并行计算可以显著地提高程序的执行效率,具有实际意义。