최근 위성항법시스템(GNSS, Global Navigation Satellite System) 기반 측위 기술의 활용 범위가 확대되면서, 다양한 분야에서 보다 향상된 정확도의 사용자 위치 산출에 대한 요구가 증가하고 있다. 현재 차량 내비게이션, 스마트폰 등에서 사용되고 있는 코드의사거리 관측치 기반의 측위 정확도는 수 미터 수준으로, 이를 극복하기 위해서는 반송파위상 관측치 기반의 측위 기술의 활용성을 높이는 것이 필요하다. 그러나 기존의 가장 대표적인 반송파위상 관측치 기반 측위 기술인 RTK(Real-Time-Kinematic) 방식은 일반적으로 이중주파수 신호 처리가 가능한 고가의 GNSS 수신기를 필요로 한다. 뿐만 아니라 높은 측위 성능 유지를 위해 사용자의 수 킬로미터 이내에 기준국이 위치해야 한다는 제약이 있다. 따라서 반송파위상 관측치 기반 측위 기술의 활용 범위 확대를 위해서는 중·장기선에서 활용 가능한 단일주파수 기반의 RTK 측위 기술 개발이 필요하다.
이에 본 논문의 연구에서는 다중 위성항법시스템을 이용한 단일주파수 기반 RTK 측위 알고리즘을 개발하였다. 측위 알고리즘의 강건성을 향상시키기 위해 각종 측위 오차 요소를 보정하는 과정에서 외부로부터 별도의 오차 보정정보를 수신하지 않고 사용자 수신기에서 모델링하여 적용하였다. 또한 충분한 수의 가시 위성 확보를 통한 다양한 사용자 관측 환경에서의 측위 성능 확보를 위하여 다중 위성항법시스템을 활용하였다. 그리고 해당 측위 알고리즘을 중·장기선에 적용하여 그 성능을 검증하고, 활용 범위 확대 가능성을 평가하였다. 이때 측위 성능 평가를 위해 최초 10분 이후의 측위 결과를 수치 및 그래프로 제시하였으며, 30cm의 3차원 위치 정확도를 기준으로 실험 결과를 분석하였다.
본 연구에서 개발한 측위 알고리즘의 성능 검증을 위해 기선거리 약 10m의 단기선 실험을 선행하였다. 1시간 GNSS 관측 데이터를 이용한 단기선 실험을 통해 GPS(Global Positioning System), BDS(Beidou Satellite System), 그리고 GLONASS(GLObal NAvigation Satellite System) 이상 3개 위성항법시스템 통합 적용 시 약 1-2cm 수준의 3차원 위치 정확도를 확보할 수 있었다. 그리고 해당 알고리즘을 기선거리 약 45-466km 범위 5개 기선 조합의 데이터를 이용한 중·장기선 실험에 적용하였다. 이때 전리층 및 대류권 오차를 보정하지 않았을 때의 측위 결과를 통해 기선거리 증가에 따른 해당 오차 성분의 영향을 확인하였다. 또한 전리층 및 대류권 오차를 보정하여 실험 대상 5개 기선에 대한 3시간 GNSS 관측 데이터 GPS/BDS/GLONASS 통합 측위를 수행한 결과, 실험 전반에서 수 센티미터 수준의 최종 측위 정확도와 최대 20cm 이내의 3차원 RMSE를 산출하였다. 그러나 실험 대상 중 최장기선(기선거리 약 466km) 실험에서는 최초 10분 이후에도 일정 시간 30cm의 목표 위치 정확도를 초과하는 경우가 발생하기도 하였다.
다양한 임계고도각 적용을 통해 사용자의 관측 환경 변화를 가정하여 위성 신호 차폐가 발생하는 환경에서의 가용성을 평가하였다. 그리고 각 조건에서 GPS 단독, GPS/BDS 통합, GPS/BDS/GLONASS 통합 측위 실험을 각각 수행하고 그 결과를 통해 다중 위성항법시스템 통합 적용에 의한 성능 향상 효과를 검증하였다. 임계고도각 30° 이하의 실험에서는 기선거리가 증가함에 따라 RMSE 수치도 대체적으로 증가하는 경향이 나타났으나, 대부분의 결과에서 본 연구에서 목표한 수준의 측위 정확도를 확보한 것을 확인할 수 있었다. 반면 임계고도각 45° 실험의 GPS 단독 적용 측위 결과에서는 기선거리 변화에 따른 경향성이 사실상 나타나지 않았으며, 본 연구의 목표 측위 성능을 확보하지 못하였다. 그러나 다중 위성항법시스템 통합 적용 측위 결과에서는 최장기선 실험을 제외하고 30cm 이내의 3차원 위치 정확도를 확보하였다. 특히 수평 성분 측위 결과는 모든 통합 측위 실험에서 최대 10cm 내외의 높은 위치 정확도를 확보할 수 있었다.
본 연구의 분석 결과를 종합했을 때, 별도의 외부 보정정보를 적용하지 않고도 단일주파수 기반의 RTK 측위로 약 10-20cm 수준의 측위 수행이 가능한 것을 확인하였다. 이를 통해 약 20-30cm 수준의 3차원 위치 정확도 혹은 10cm 내외의 수평 위치 정확도를 필요로 하는 분야에서 타 기술을 대체할 수 있을 것으로 판단된다. 뿐만 아니라 별도의 측위 오차 보정정보가 불필요하고, 약 500km 이내의 중·장기선에 대해 다중 위성항법시스템 활용이 가능하다는 사항을 고려할 때, RTK 측위 기술의 상용화에 다양하게 기여할 수 있을 것으로 기대된다.
Abstract
Recently, as the utilization of positioning technology based on Global Navigation Satellite System (GNSS) has been expanded, there is a growing demand for more accurate measurement of user location in various fields. The positioning accuracy based on code pseudo-range measurements in the car navigation system and the smartphone is several meters. In order to improve the accuracy, it is necessary to improve the usability of positioning technique based on carrier-phase measurements. However, RTK (Real-Time-Kinematic), which is the most typical carrier-phase based positioning technique, generally requires an expensive GNSS receiver capable of dual frequency signal processing. In addition, there is a restriction that the reference station should be located within a few kilometers of the user to maintain high positioning performance. Therefore, it is essential to develop a single-frequency based RTK positioning algorithm that can be used in medium/long-range baselines in order to expand the application range of carrier-phase based positioning technology.
In this paper, we developed a single-frequency based RTK positioning algorithm using a multi-constellation GNSS. In the process of correcting various positioning error factors, the user receiver corrects the error by modeling without any additional error correction information from the outside to improve the robustness of the positioning algorithm. In addition, multi-constellation GNSS is utilized to maintain positioning performance in various user observation environments by securing a sufficient number of visible satellites. Then, it was applied in medium/long-range baselines, and its performance was verified. In order to evaluate the positioning performance, numerical and graphical representations of the results after the first 10 minutes are presented, and the experimental results are analyzed based on the 3-dimensional position accuracy of 30cm. In order to verify the performance of the positioning algorithm developed in this study, a short-range baseline test with the distance of about 10m was performed. One hour GNSS observation data were used for short-range baseline experiments, and it was possible to obtain 3-dimensional position accuracy of about 1-2cm when GPS(Global Positioning System)/BDS(Beidou Satellite System)/GLONASS(GLObal NAvigation Satellite System) was integrated. And we applied it to medium/long-range baselines experiments using data from 5 baseline ranging from about 45 to 466 km. At this test, the influence of the ionospheric and tropospheric errors according to the increase of baseline distance was confirmed through the positioning results when the corresponding error component were not corrected. After that, GPS/BDS/GLONASS integrated positioning was performed for 3 hours GNSS observation data for 5 baselines of experiment by correcting ionospheric and tropospheric error. As a result, the final positioning accuracy of several centimeters and the 3- dimensional RMSE of up to 20 cm were calculated in the overall experiment. However, the longest baseline(the distance of about 466km) of the experiment also exceeded the numerical goal of position accuracy of 30cm after the first 10 minutes.
We evaluated the availability in the environment where the satellite signal blocked occurs by assuming the change of the user''s observation environment by applying various elevation cutoffs. In addition, GPS/BDS and GPS/BDS/GLONASS integrated positioning experiments were performed under each condition, and the performance enhancement was verified by integrating multiple satellite navigation systems. RMSE values increased with increasing baseline distance at 30° or less of elevation cutoff. However, most of the results show that the positioning accuracy is achieved in the numerical goal of this study. On the other hand, the result of GPS-only processing at 45° elevation cutoff did not show the tendency according to the change of baseline distance, and the target positioning performance of this study was not secured. However, in the result of multi-constellation GNSS integrated positioning, 3-dimensional positioning accuracy within 30cm was secured except for the longest baseline experiment. In particular, horizontal component positioning results showed high position accuracy of up to about 10cm in all integrated positioning tests.
Based on the analysis results of this study, it is confirmed that it is possible to perform positioning of about 10-20cm level by single- frequency based RTK positioning without applying additional error correction. It is expected that it will be possible to replace other technologies in the field which requires 3-dimensional position accuracy of about 20-30cm or horizontal position accuracy of about 10cm. In addition, it is also expected that it will contribute variously to the generalization of RTK positioning technology considering the fact that additional positioning error correction information is unnecessary and multi-constellation GNSS can be used for medium/long-range baselines within about 500km.
목차
목 차제 1 장 서 론 ·········································································· 011.1 연구 배경 및 필요성 ························································· 011.2 국내·외 연구 동향 ····························································· 051.3 연구 목적 및 내용 ····························································· 07제 2 장 위성항법시스템 개요 ······················································ 092.1 위성항법시스템 구성 ························································· 092.2 위성항법시스템 관측치 ······················································ 102.3 위성항법시스템 측위 개념 ·················································· 122.3.1 코드의사거리 기반 측위 모델 ······································ 122.3.2 반송파위상 기반 측위 모델 ········································· 132.4 측위 오차 요소 ································································· 14제 3 장 반송파 위상 기반 이중차분 측위 알고리즘 설계 ················· 173.1 상대측위와 이중차분 ························································· 173.2 측위 오차 모델링 ······························································ 223.2.1 전리층 오차 ······························································ 243.2.2 대류권 오차 ······························································ 253.2.3 위성 궤도 오차 ························································· 263.2.4 신호 전달 시간 및 지구 자전 효과 ······························· 273.2.5 위성 시계 오차 및 상대성 효과 ··································· 283.2.6 사이클슬립 ································································ 313.2.7 반송파위상 모호정수 ··················································· 323.3 다중 위성항법시스템 통합 ·················································· 333.3.1 위성 신호 주파수 ······················································· 353.3.2 좌표계 및 시간계 ······················································· 363.3.3 위성 궤도 결정 방식 ··················································· 383.4 측위 알고리즘 설계 ··························································· 403.4.1 확장 칼만 필터 적용 ··················································· 403.4.2 미지 변수 초기값 결정 ··············································· 443.4.3 다중 위성항법시스템 통합 RTK 측위 모델 행렬 구성 ····· 45제 4 장 측위 알고리즘 중·장기선 적용 성능 평가 ·························· 494.1 단기선 실험을 통한 측위 알고리즘 성능 검증 ······················· 494.2 중·장기선 실험 데이터 수집 ··············································· 544.3 측위 알고리즘의 중·장기선 적용 실험 결과 및 분석 ·············· 574.3.1 측위 알고리즘의 중·장기선 적용 성능 분석 ···················· 574.3.2 다중 위성항법시스템 통합 적용 효과 검증 ····················· 64제 5 장 결 론 ·········································································· 70참고문헌 ····················································································· 72