基于FGO的GNSS/INS组合导航完好性监测算法  

Integrity Monitoring for GNSS/INS Integration Based on Factor Graph Optimization

作　　者：兰晓伟 许承东[1] 赵靖 LAN Xiao-wei;XU Cheng-dong;ZHAO Jing(School of Aerospace Engineering,Beijing Institute of Technology,Beijing 100081,China;China Transport Telecommunications and Information Center,Beijing 100011,China)

机构地区：[1]北京理工大学宇航学院,北京100081 [2]中国交通通信信息中心,北京100011

出　　处：《计算机仿真》2023年第9期7-12,共6页Computer Simulation

摘　　要：针对GNSS/INS组合导航中GNSS卫星的故障风险,提出一种基于因子图优化的完好性监测算法。通过建立GNSS/INS组合导航因子图模型和各类因子的观测函数,将组合导航的非线性数据融合问题转化为多次迭代的线性最小二乘问题,利用线性最小二乘残差构造检验统计量进行故障检测,并通过定义斜率寻找最大化斜率的最坏故障情况,基于最坏故障情况实现保护级计算。动态仿真结果表明,相较于传统算法,所设计的组合导航完好性监测算法故障检测率达到100%的伪距故障偏差由45m减小到15m,并且提供的保护级在有效包络漏检情况下的定位误差的前提下降低约10m。To protect the user from GNSS satellite fault risk in the GNSS/INS integration,an integrity monitoring method based on factor graph optimization is proposed.By establishing the factor graph model of the GNSS/INS integration and the observation functions of various types of factors,the nonlinear data fusion problem of GNSS/INS integration was transformed into the linear least squares problem with multiple iterations.And the linear least squares residuals were utilized to construct test statistics for fault detection.The protection level was calculated based on the worst fault case which maximizes the defined slope.Simulation experiments involving dynamic data show that,compared with the traditional algorithm,the designed integrity monitoring method reduces the pseudo range bias from 45m to 15m when the fault detection rate reaches 100%.Moreover,the provided protection level,which can effectively bound the positioning error even under the missed detection condition,is reduced by about 1Om.

关 键 词：完好性监测 因子图优化 故障检测 保护级 

分 类 号：TN967.2[电子电信—信号与信息处理]