{"title":"基于机器学习的空间域多路径建模在GNSS短基线处理中的应用","authors":"Yuanxin Pan, Gregor Möller, Benedikt Soja","doi":"10.1007/s10291-023-01553-y","DOIUrl":null,"url":null,"abstract":"Abstract Multipath is the main unmodeled error source hindering high-precision Global Navigation Satellite System data processing. Conventional multipath mitigation methods, such as sidereal filtering (SF) and multipath hemispherical map (MHM), have certain disadvantages: They are either not easy to use or not effective enough for multipath mitigation. In this study, we propose a machine learning (ML)-based multipath mitigation method. Multipath modeling was formulated as a regression task, and the multipath errors were fitted with respect to azimuth and elevation in the spatial domain. We collected 30 days of 1 Hz GPS data to validate the proposed method. In total, five short baselines were formed and multipath errors were extracted from the postfit residuals. ML-based multipath models, as well as observation-domain SF and MHM models, were constructed using 5 days of residuals before the target day and later applied for multipath correction. It was found that the XGBoost (XGB) method outperformed SF and MHM. It achieved the highest residual reduction rates, which were 24.9%, 36.2%, 25.5% and 20.4% for GPS P1, P2, L1 and L2 observations, respectively. After applying the XGB-based multipath corrections, kinematic positioning precisions of 1.6 mm, 1.9 mm and 4.5 mm could be achieved in east, north and up components, respectively, corresponding to 20.0%, 17.4% and 16.7% improvements compared to the original solutions. The effectiveness of the ML-based multipath model was further validated using 30 s sampling data and data from a low-cost device. We conclude that the ML-based multipath mitigation method is effective, easy to use, and can be easily extended by adding auxiliary input features, such as signal-to-noise ratio, during model training.","PeriodicalId":12788,"journal":{"name":"GPS Solutions","volume":null,"pages":null},"PeriodicalIF":4.5000,"publicationDate":"2023-10-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":"{\"title\":\"Machine learning-based multipath modeling in spatial domain applied to GNSS short baseline processing\",\"authors\":\"Yuanxin Pan, Gregor Möller, Benedikt Soja\",\"doi\":\"10.1007/s10291-023-01553-y\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"Abstract Multipath is the main unmodeled error source hindering high-precision Global Navigation Satellite System data processing. Conventional multipath mitigation methods, such as sidereal filtering (SF) and multipath hemispherical map (MHM), have certain disadvantages: They are either not easy to use or not effective enough for multipath mitigation. In this study, we propose a machine learning (ML)-based multipath mitigation method. Multipath modeling was formulated as a regression task, and the multipath errors were fitted with respect to azimuth and elevation in the spatial domain. We collected 30 days of 1 Hz GPS data to validate the proposed method. In total, five short baselines were formed and multipath errors were extracted from the postfit residuals. ML-based multipath models, as well as observation-domain SF and MHM models, were constructed using 5 days of residuals before the target day and later applied for multipath correction. It was found that the XGBoost (XGB) method outperformed SF and MHM. It achieved the highest residual reduction rates, which were 24.9%, 36.2%, 25.5% and 20.4% for GPS P1, P2, L1 and L2 observations, respectively. After applying the XGB-based multipath corrections, kinematic positioning precisions of 1.6 mm, 1.9 mm and 4.5 mm could be achieved in east, north and up components, respectively, corresponding to 20.0%, 17.4% and 16.7% improvements compared to the original solutions. The effectiveness of the ML-based multipath model was further validated using 30 s sampling data and data from a low-cost device. We conclude that the ML-based multipath mitigation method is effective, easy to use, and can be easily extended by adding auxiliary input features, such as signal-to-noise ratio, during model training.\",\"PeriodicalId\":12788,\"journal\":{\"name\":\"GPS Solutions\",\"volume\":null,\"pages\":null},\"PeriodicalIF\":4.5000,\"publicationDate\":\"2023-10-16\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"0\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"GPS Solutions\",\"FirstCategoryId\":\"1085\",\"ListUrlMain\":\"https://doi.org/10.1007/s10291-023-01553-y\",\"RegionNum\":1,\"RegionCategory\":\"地球科学\",\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"Q1\",\"JCRName\":\"REMOTE SENSING\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"GPS Solutions","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1007/s10291-023-01553-y","RegionNum":1,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q1","JCRName":"REMOTE SENSING","Score":null,"Total":0}
Machine learning-based multipath modeling in spatial domain applied to GNSS short baseline processing
Abstract Multipath is the main unmodeled error source hindering high-precision Global Navigation Satellite System data processing. Conventional multipath mitigation methods, such as sidereal filtering (SF) and multipath hemispherical map (MHM), have certain disadvantages: They are either not easy to use or not effective enough for multipath mitigation. In this study, we propose a machine learning (ML)-based multipath mitigation method. Multipath modeling was formulated as a regression task, and the multipath errors were fitted with respect to azimuth and elevation in the spatial domain. We collected 30 days of 1 Hz GPS data to validate the proposed method. In total, five short baselines were formed and multipath errors were extracted from the postfit residuals. ML-based multipath models, as well as observation-domain SF and MHM models, were constructed using 5 days of residuals before the target day and later applied for multipath correction. It was found that the XGBoost (XGB) method outperformed SF and MHM. It achieved the highest residual reduction rates, which were 24.9%, 36.2%, 25.5% and 20.4% for GPS P1, P2, L1 and L2 observations, respectively. After applying the XGB-based multipath corrections, kinematic positioning precisions of 1.6 mm, 1.9 mm and 4.5 mm could be achieved in east, north and up components, respectively, corresponding to 20.0%, 17.4% and 16.7% improvements compared to the original solutions. The effectiveness of the ML-based multipath model was further validated using 30 s sampling data and data from a low-cost device. We conclude that the ML-based multipath mitigation method is effective, easy to use, and can be easily extended by adding auxiliary input features, such as signal-to-noise ratio, during model training.
期刊介绍:
GPS Solutions is a scientific journal. It is published quarterly and features system design issues and a full range of current and emerging applications of global navigation satellite systems (GNSS) such as GPS, GLONASS, Galileo, BeiDou, local systems, and augmentations. Novel, innovative, or highly demanding uses are of prime interest. Areas of application include: aviation, surveying and mapping, forestry and agriculture, maritime and waterway navigation, public transportation, time and frequency comparisons and dissemination, space and satellite operations, law enforcement and public safety, communications, meteorology and atmospheric science, geosciences, monitoring global change, technology and engineering, GIS, geodesy, and others.
GPS Solutions addresses the latest developments in GNSS infrastructure, mathematical modeling, algorithmic developments and data analysis, user hardware, and general issues that impact the user community. Contributions from the entire spectrum of GNSS professionals are represented, including university researchers, scientists from government laboratories, receiver industry and other commercial developers, public officials, and business leaders.