GNSS惯导紧组合中改进的容积卡尔曼滤波算法  

The improved cabature Kalman filter in GNSS/INS tightly coupled mode

作　　者：徐爱功[1] 邹鑫慈 袁庆[2] 房穹 隋心[1] XU Aigong;ZOU Xinci;YUAN Qing;FANG Qiong;SUI Xin(School of Surveying and Geography Science,Liaoning Technical University,Fuxin,Liaoning 123000,China;Engineering Survey Technology Application Research Institute,China Railway Siyuan Survey and Design Group Co.,LTD.,Wuhan 430063,China)

机构地区：[1]辽宁工程技术大学测绘与地理科学学院,辽宁阜新123000 [2]中铁第四勘察设计院集团有限公司工程测量技术应用研究所,武汉430063

出　　处：《测绘科学》2022年第3期22-28,共7页Science of Surveying and Mapping

基　　金：辽宁省重点研发计划项目(2020JH2/10100044);国家自然科学基金项目(42074012);辽宁省自然科学基金计划指导计划项目(2019-ZD-0051);辽宁省教育厅基础研究项目(LJ2020JCL016)。

摘　　要：针对常规容积卡尔曼滤波在GNSS/INS组合导航系统理论模型与实际模型不匹配情况下滤波精度下降甚至发散的问题,该文将强跟踪滤波理论与容积卡尔曼滤波相结合形成STCKF算法,并引入时变噪声估计器。在滤波过程中,该算法利用次优渐消因子调整预测状态误差协方差,实时调整增益矩阵进而保证滤波器对实际系统状态的跟踪,并通过时变噪声估计器对量测噪声进行修正。实验结果表明,所提算法提高了常规非线性滤波器对不确定系统模型的鲁棒性及跟踪系统状态突变的能力,有效抑制异常观测信息对导航解的影响,提升了导航系统的精度和稳定性。Aiming at the problem of conventional cubature Kalman filter reduced accuracy or even divergence under the situation where the theoretical model of system did not match actual model.This paper combined the strong tracking theory with the cubature Kalman filter to form the STCKF algorithm,and a time-varying noise estimator was be introduced.In the filtering process,the predicted state error covariance was to be adjusted by suboptimal fading factor,then the gain matrix was adjusted online to ensure that the filter tracks the actual system state,and the time-variable measurement noise was revised by robustness factor.The experimental results showed that the proposed algorithm improved the robustness of the conventional cubature Kalman filter to uncertain system model and the ability to track system state mutation,effectively suppressed the influence of abnormal observation information on the navigation solution and enhanced the accuracy and stability of the navigation system.

关 键 词：容积卡尔曼滤波 时变噪声估计 强跟踪滤波 渐消因子 

分 类 号：P228[天文地球—大地测量学与测量工程]