摘要
针对GNSS伪距相对定位中多路径效应和非视距NLOS误差导致定位精度降低的问题,该文提出了一种基于多粗差探测与抗差估计联合处理多路径/NLOS误差的方法。该方法以“由粗到细、逐步精化”的基本原则,首先,构建伪距相对定位双差方程,基于最小二乘获取各观测值改正数及对应权阵;在此基础上,建立检验方差因子统计量,并结合检验限值迭代进行多粗差探测,初步剔除大部分含多路径/NLOS粗差的观测值;最后,引入抗差估计理论,根据剩余观测量的标准化残差重新评估各观测值权重,通过选权迭代实现对包含多路径/NLOS误差观测量的剔除与降权,提升整体定位效果。实验结果表明,单独应用多粗差探测法或抗差估计法在处理多路径和NLOS误差时,其适用性存在限制。本文提出的联合处理方法有效整合了两种方法的优点,显著提升了误差检测的灵敏度和算法的稳健性,在不同测试环境中均表现出了较高的性能和可靠性。
Aiming at the problem that multipath effect and NLOS error lead to the decrease of positioning accuracy in GNSS pseudo-range relative positioning,a method based on multi-gross errors detection and robust estimation is proposed to deal with multipath/NLOS error in this article.The method is based on the basic principle of“from coarse to fine,step by step refinement.”Firstly,the double difference equation of pseudo-range relative positioning is constructed,and the correction of each observation value and the corresponding weight matrix are obtained based on the least square method;on the basis of which,the test variance factor statistic is established,and the multi-gross errors are detected by combining the test limit iteration,most of the observations with multipath/NLOS gross errors are preliminarily eliminated;finally,the theory of robust estimation is introduced,and the weights of each observation are reassessed based on the standardized residuals of the residual observations.Through weight selection iteration,observations containing multipath/NLOS errors are eliminated or down-weighted,thereby improving the overall positioning performance.The experimental results show that the applicability of the multi-gross errors detection method or the robust estimation method is limited when dealing with multipath and NLOS errors.The joint processing method proposed in this paper effectively integrates the advantages of the two methods,significantly improves the sensitivity of error detection and the robustness of the algorithm,and shows high performance and reliability in different test environments.
作者
魏书恩
邓健
王华银
任春华
WEI Shuen;DENG Jian;WANG Huayin;REN Chunhua(School of Computer and Information Engineering,Xiamen University of Technology,Xiamen,Fujian 361024,China)
出处
《测绘科学》
北大核心
2025年第6期71-79,共9页
Science of Surveying and Mapping
基金
福建省自然科学基金项目(2022J011237,2023J011431)。
关键词
伪距相对定位
多路径
NLOS误差
多粗差探测
抗差估计
pseudo-range relativepositioning
multi-path
NLOS error
1 multi-gross errors detection
robust estimation
