基于CEEMDAN的多路径实时修正模型研究
Research on Multipath Real-time Correction Model Based on CEEMDAN
孙博文 1朱星盛 2杨怀志 2匡团结 1张云龙 1刘洪润 2肖翔2
作者信息
- 1. 中国铁路设计集团有限公司,天津 300308;天津市轨道交通导航定位及时空大数据技术重点实验室,天津 300308
- 2. 京沪高速铁路股份有限公司,北京 100038
- 折叠
摘要
在北斗变形监测应用中,多路径误差显著影响监测精度.传统方法如小波分析和经验模态分解(EMD)在提取多路径误差信号过程中,常受模态混叠等问题影响,限制了多路径误差改正精度的提升.将完全自适应噪声集合经验模态分解算法(CEEMDAN)引入北斗监测,实现多路径误差信号的有效提取,并借助半天球模型对多路径误差进行实时改正.实验结果显示,在DOY(123～129)期间,该模型显著提高坐标序列精度:均方根误差(RMS)在水平北向由0.83 cm改善至0.09 cm,水平东向由1.05 cm改善至0.13 cm,高程方向由7.27 cm改善至0.35 cm.此外,通过利用7天的半天球模型内插DOY 130的数据,验证了该方法的实时性和有效性,实现水平精度优于0.15 cm、高程精度优于0.35 cm的改正效果,RMS改善率在水平和高程方向分别优于93.5％和95.5％.
Abstract
In the application of Beidou deformation monitoring,multipath error significantly affects the monitoring accuracy.Traditional methods such as wavelet analysis and empirical mode decomposition(EMD)are often affected by modal aliasing in the process of extracting multipath error signals,which limits the improvement of multipath error correction accuracy.The complete ensemble empirical mode decomposition with adaptive noise(CEEMDAN)was introduced into Beidou monitoring to extract multipath error signals effectively,and the hemispherical model was used for real-time correction of multipath errors.The experimental results show that the model significantly improves the accuracy of coordinate sequence during DOY(123～129):the root mean square error(RMS)was improved from 0.83 cm to 0.09 cm in horizontal north direction,from 1.05 cm to 0.13 cm in horizontal east direction and from 7.27 cm to 0.35 cm in vertical direction.In addition,the real-time performance and effectiveness of this method were verified by interpolating DOY 130 data with a 7-day hemispherical model.The corrected horizontal accuracy was better than 0.15 cm,the vertical accuracy was better than 0.35 cm,and the RMS improvement rates in the horizontal and vertical directions were higher than 93.5％and 95.5％respectively.
关键词
京沪高铁/变形监测/多路径误差/CEEMDAN/半天球模型/实时修正/北斗系统Key words
Beijing-Shanghai HSR/deformation monitoring/multipath error/CEEMDAN/hemispherical model/real-time correction/beidou system引用本文复制引用
基金项目
中国国家铁路集团有限公司科技研发计划(P2022X001)
天津市自然科学基金重点项目(23JCZDJC00670)
内蒙古自治区重点研发和成果转化计划(2023SKJHZ0269)
中国铁路设计集团有限公司科研开发项目(2023A0240102)
中国铁路设计集团有限公司科研开发项目(2023A0253801)
中国铁路设计集团有限公司科研开发项目(2023A0240105)
出版年
2024
NETL
NSTL
万方数据