RAE并行滤波的重力异常匹配算法  

A Matching Algorithm Using Gravity Anomaly Based on the RAE Parallel Filtering

作　　者：王伟[1,2] 李姗姗 邢志斌[1] 李中全[3] 黄韵颖[3] 

机构地区：[1]信息工程大学地理空间信息学院,河南郑州450001 [2]63666部队,新疆维吾尔自治区巴音郭楞蒙古自治州841000 [3]78138部队,四川成都610000

出　　处：《武汉大学学报（信息科学版）》2016年第12期1664-1670,共7页Geomatics and Information Science of Wuhan University

基　　金：国家863计划(2013AA122502);国家自然科学基金(41274029)~~

摘　　要：扩展卡尔曼滤波技术可以利用序列观测重力异常数据以及航行区域重力异常基准图来校正惯性导航系统漂移误差。针对因重力测量环境的变化、测量仪器扰动等因素造成的重力异常观测噪声不确定问题,提出了基于量测残差自适应估计观测噪声协方差(residual-based adaptive estimation,RAE)的重力异常滤波匹配算法;设计了一组并行卡尔曼滤波器,并简化了最优滤波器的选择准则。不同重力特征区域的实验表明,该算法能够有效降低惯性导航系统经纬向漂移误差,提高系统的导航定位精度。Serial observed gravity anomaly data and a gravity anomaly referenced map for navigation can be used to correct the drifting errors of inertial navigation system based on the EKF.To address the problem of unknown gravity anomaly measurement noise due to an unpredictable gravimetric environment and disturbances to the measuring instruments,et al,a matching algorithm for gravity anomaly filtering based on residual errors can be used to estimate measurement noise variance adaptively;Residual-based Adaptive Estimation（RAE）.A set of parallel Kalman filters were designed and a rule for selecting the best filter was simplified.RAE filtering experimental results show that the longitude and latitude drifting errors in inertial navigation systems can be reduced effectively based on the RAE filtering and positioning accuracy of the navigation system thus improved.

关 键 词：RAE滤波 重力异常 并行卡尔曼滤波 自适应估计 观测噪声 

分 类 号：P223[天文地球—大地测量学与测量工程]