摘要
针对现有惯性/重力/重力梯度组合导航EKF匹配算法滤波状态方程存在模型误差以及惯性导航系统定位存在累积误差而造成的滤波失准乃至发散问题,提出一种基于CEP(Circular Error Probable,圆概率误差)的自适应并行EKF匹配算法.该算法首先通过自适应因子调节状态预测信息的权重,削弱预设动力学模型不准确产生的误差;同时利用惯性导航系统圆概率误差半径构建基于CEP的移动窗口分层模型,然后根据滤波量测值与窗口坡度等信息,对移动窗口范围进行约束;最后组建分层窗口并行滤波器,得到最优匹配结果.南海海域实验结果表明,基于CEP的自适应并行EKF匹配算法相较于传统EKF算法和自适应EKF算法的水下重力匹配导航定位精度分别提升了74.0%和49.8%.该算法能够在一定程度上克服惯性导航系统由于时间推移误差积累的缺陷,提高系统导航定位精度,增加匹配算法的鲁棒性.
In the existing inertial,gravity and gravity gradient integrated matching navigation,the model error of the EKF state equation and the filtering inaccuracy and divergence caused by the cumulative error of the inertial navigation system(INS)is of critical importance,an adaptive parallel EKF matching algorithm based on CEP(circular error possible)is proposed in this paper.Firstly,to weaken the error caused by the inaccuracy of the preset dynamic model,the weight of state prediction information is adjusted by adaptive factors.And using the circular probability error radius of INS,the sliding window layered model based on CEP is constructed simultaneously.Then,based on the filter measurement and window slope,the range of sliding window is constrained.Finally,a layered window parallel filter is established to obtain the optimal result.The experimental results in the South China Sea show that compared with the traditional EKF and the adaptive EKF algorithm,the positioning accuracies of proposed method are improved by 74.0%and 49.8%,respectively.With the proposed algorithm,the accumulation of time-lapse errors using INS can be shortened to some extent,and the positioning accuracy and robustness can be enhanced.
作者
黄炎
李姗姗
范雕
谭勖立
冯进凯
吕明昊
HUANG Yan;LI ShanShan;FAN Diao;TAN XuLi;FENG JinKai;LV MingHao(National Innovation Institute of Defense Technology,Academy of Military Science,Beijing 100071,China;PLA Strategic Support Force Information Engineering University,Zhengzhou 450001,China;96941 Troops,Beijing 100080,China)
出处
《地球物理学报》
SCIE
EI
CAS
CSCD
北大核心
2024年第1期50-62,共13页
Chinese Journal of Geophysics
基金
国家自然科学基金面上项目(42174007,42174013)资助。
