航空工程进展2025，Vol.16Issue(5)：58-68,11.DOI:10.16615/j.cnki.1674-8190.2025.05.06
基于K-LSTM模型的卫星定位误差估计方法
Positioning error estimation method based on K-LSTM model
摘要
Abstract
With the development of Global Navigation Satellite System(GNSS),satellite-based positioning tech-nology has become an important data source for aviation navigation.However,in scenarios involving unmanned ur-ban air mobility(UAM)applications,satellite positioning is susceptible to multipath(MP)and non line of sight(NLOS)signals leading to deterioration in positioning accuracy,posing a challenge to aircraft safety.To address this problem,a method utilizing the K-LSTM model for satellite positioning error estimation is proposed.Firstly,the K-means clustering method is used to detect MP/NLOS signals.Secondly,the relationship between satellite observations and positioning errors in different environments is investigated and the neural network model is extend-ed.This extension involves adding a droupout layer,a ReLU layer,a fully-connected layer,and a regression layer on top of the Long Short-Term Memory(LSTM)neural network.Finally,the extended LSTM model is used to estimate and correct the localization error caused by MP/NLOS signals.The experimental results show that in stat-ic urban canyon and dynamic ground reflection environments,the positioning errors in the east,north,and up direc-tions of MP/NLOS signals corrected by the extended LSTM model are significantly reduced compared to before corrections,and the positioning accuracy is significantly improved.关键词
卫星定位/K-means聚类/多径和非视距/扩展长短时记忆神经网络模型/误差校正Key words
satellite positioning/K-means clustering/multipath and non line of sight/extended long short-term memory model/error correction分类
航空航天引用本文复制引用
刘瑞华,刘志阳,马赞,郑智明,钟科林..基于K-LSTM模型的卫星定位误差估计方法[J].航空工程进展,2025,16(5):58-68,11.基金项目
国家重点研发计划(2022YFB3904304) （2022YFB3904304）
