【作者】 李姗姗；

【导师】 吴晓平；

【作者基本信息】 解放军信息工程大学 ， 大地测量学与测量工程， 2010， 博士

【摘要】 本文主要针对水下重力无源导航系统的特点,研究无源重力辅助惯性导航的基本理论与方法,利用各种重力特征传感器获取信息,进行重力图重构、数据预处理、匹配算法分析、性能评估,然后进行智能化数据融合,进而完成水下无源导航模块化的仿真设计,以便能有效地修正惯性导航误差,从而获得较高的定位精度,为真实海洋环境下水下潜器的精确导航提供可靠的技术依据。作者在文中所作的主要工作和创新点概括如下:1.归纳总结了水下重力无源辅助惯性导航系统的基本构架;论述了平台式惯性导航系统的相关基本理论;给出了卫星测高的基本原理以及基于卫星测高数据建立海洋重力异常图的理论与方法;介绍了重力与重力梯度传感器测量系统;综述了各种经典匹配算法的工作原理及其特点。2.在现有海洋2′×2′重力图数据的水平下,提出了基于孔斯曲面的海洋重力异常图重构方法,建立了双一次和双三次孔斯曲面重力异常模型,较之于移动曲面拟合,精度均得到了有效提高;研究了基于30"×30"海深信息构建高精度、高分辨率重力基准图的方法。3.研究了水下实时重力测量的可行性及其基于INS输出的重力扰动改正的实现途径,分析了水下重力传感器测量误差的物理特性并通过相应的数学模型探讨了其量级大小;研究探讨了重力传感器数据的预处理、延拓归算及数学转换问题。4.研究分析了海洋重力场统计特征对重力无源匹配导航精度的影响;基于利用重力信息对惯性系统进行修正主要取决于重力异常空间属性变化特征的思想,进行了局部重力场特征参数的统计计算分析;通过不同海域的重力仿真匹配实验,探讨了重力适配区的问题。5.针对辅助导航明确的研究对象-惯性导航系统,研究探讨了惯性导航陀螺漂移、加速度零位误差、初始校准误差等主要误差源的噪声传播特性;建立了符合惯性导航特点的航线误差仿真模型;根据东向位置误差的无界性特点,提出了匹配结果应更关注东向位置修正精度的建议。6.为了削弱重力测量中厄特弗斯效应、正常重力等计算受惯导误差的影响,研究提出了以相邻观测重力异常之差组成新的观测序列进行相关匹配的差分降相关极值算法;提出了由于受重力测量、重力图、惯导等随机干扰误差影响导致源于正确位置出现多个有效位置的概率数据关联滤波算法,提高了匹配算法的可用性与鲁棒性。7.研究了基于费波纳奇数列调优的ICCP匹配算法;针对ICCP匹配算法的局限性,提出了先基于序列相关极值匹配算法将惯导航迹修正到真实航迹附近,然后将得到的匹配航迹作为惯导航迹再进行ICCP匹配,从而将获得的最优航迹作为最终匹配结果的融合算法。8.在SITAN算法的基础上,研究了双二次曲面的重力异常随机线性化技术;通过对重力异常量测方程各项量级的探讨,建立了优化的量测方程;研究了并行卡尔曼滤波技术;从卡尔曼滤波稳定性的角度探讨了为满足可观性矩阵可逆的采样周期的选取问题。9.研究设计了将基于惯性导航力学编排方程的扰动重力与垂线偏差的开环补偿与重力信息匹配以修正惯性导航随机误差的闭环补偿的集成重力梯度辅助惯性导航系统的工作模式;初步探讨了无图模式下多种重力场元联合匹配的方法。10.研究推导了扰动重力矢量对惯性导航系统位置与速度的误差传播模型,分析了扰动重力矢量引起的惯性导航系统位置与速度误差量级的大小及特性,定量、定性地验证了重力开环补偿的必要性。11.通过实验计算证明了重力梯度的频域化延拓既能满足精度要求,又不损失实时性,能满足水下重力梯度匹配导航实时性的需求;通过对梯度数据量级及其变化统计特征的分析研究,提出了利用主对角独立的两个特征变化明显的张量Tx x、T yy作为观测量的卡尔曼滤波匹配算法。12.设计了水下重力无源辅助导航的数据总体处理流程,构建了面向对象应用的水下重力无源导航系统仿真实验集成平台。更多还原

【Abstract】 Aiming at the characteristics of the automatic underwater gravity navigation system, the fundamental theories and methods are mainly studied in this dissertation. Gravity map reconstruction, data preprocessing, analysis of matching algorithms and capability evaluation are made with the information obtained by various gravitational sensors. Then, the intelligentized data is syncretized, and the simulation design of the automatic underwater navigation modules is completed so as to correct the inertial navigation errors and improve the positioning precision. Thus, it could be a reliable technology for the precise navigation of underwater carriers in the ocean. The main works and innovative ideas are listed as follows:1. The basic skeleton of the automatic underwater gravity aided inertial navigation system is summarized in this thesis, while the correlative fundamental theories are discussed, the basic principle of the Satellite Altimetry and the oceanic gravity anomaly map based on satellite altimetry data are given. The gravity and gravity gradient sensor surveying systems are introduced. What’s more, the principles and characteristics of the classical matching algorithms are summed up as well.2. With the oceanic gravity map data at the level of 2′×2′, the reconstruction method of oceanic gravity anomaly map based on coons curved surface is put forward. Besides, the gravity anomaly model of double one order or three orders coons curved surface are established. The precision are much better than that of the mobile curve surface interpolation. The method of establishing high precision and high resolution gravity datum map based on the depth of the ocean at the level of 30"×30" is studied.3. The feasibility of the real time underwater gravity measurement and the implement method of gravity disturbing correction based on the output of INS are studied. Then, the physical characteristics of the underwater gravity sensor measurement errors are analyzed and the magnitude is discussed by corresponding mathematical models. Also, the pre-process, continuation reduction as well as the mathematical transformation of the gravity sensor measurement data are discussed.4. The influence of the statistical characteristics of the oceanic gravity field to the precision of automatic gravity navigation is analyzed. Based on the thought that the correction of the navigation system mostly determined by the variable characteristics of space properties, statistical calculation analysis of local gravity field characteristic parameters is made. The gravity matching zone is discussed by simulation gravity experiments of different sea areas.5. Aimed at the inertial navigation system, which is the study object of the noise transmission characteristics of the main error sources such as gyroscopes drift in inertial navigation, zero position errors of the acceleration, initial adjustment errors and so on are discussed. A simulated model of sailing path errors that fulfilling the characteristics of inertial navigation is established. Furthermore, a suggestion that the eastern correction precision should be paid more attention to in the matching result is put forward when considering the unbounded characteristic of the eastern position errors.6. In order to eliminate the influence of the inertial errors in the calculation of E?tv?s effect and the normal gravity in gravity measurement, differential descending correlative extremum matching algorithm is put forward with new observational series comprised of the differential adjacent gravity anomaly observations. Probabilistic data association filtering algorithm is put forward with the fact that multi-available positions are resulted from the disturbing errors of the real position, and the reliability and robustness of the matching algorithm is improved.7. The ICCP matching algorithm based on Fibonacci number series is studied. As for its localization, the following strategy is taken in this thesis. With serial relevant extremum matching algorithm, the inertial flight path is corrected to close to the real one before the ICCP is taken to the matched path to obtain the best sailing path.8. With the help of SITAN, the random linearization technology of the double two orders curved surface of gravity anomaly is discussed. Optimized gravity anomaly measurement functions are constructed by the discussion of all the items magnitude. Parallel Kalman filtering technology is studied, and how to determine the sampling period for the inverse of the visible matrix is considered to ensure the stability of the Kalman filtering.9. A compositive gravity gradient aided inertial navigation system working mode is designed, which is based on the thought that the close-loop compensation of the random inertial navigation errors is corrected through the matching of gravity information and the open loop compensation of the disturbing gravity and the deflection of the vertical arranged by the inertial navigation mechanic functions. In addition, the method of variable gravity cells matching method without gravity map is discussed preliminarily.10. The error transmission model of the position and velocity in the inertial navigation system by the disturbing gravity vector is studied, of which the magnitude and characteristics are analyzed to validate the necessity of the gravity open loop compensation qualitatively and quantitatively.11. Experiments present that the frequency domain continuation of the gravity gradient could satisfy the precision as well as the real time requirements of the underwater gravity gradient matching navigation. Kalman filtering matching algorithm with independent diagonal element Txx and Tyy, which both have obvious characteristics variance, is put forward based on the magnitude and variance statistical characteristics analysis of the gradient data.12. The data processing flow of the automatic underwater gravity aided navigation is designed, of which the object-oriented compositive simulation experiments platform is established.更多还原