A GNSS Multi-Antenna Fast Millimeter-level Positioning Method for Rail Track Deformation Monitoring

Accurate rail track deformation and regularity monitoring is the basic condition to ensure safe operation of high-speed railway. The technical challenge that needs to be addressed is to achieve positioning accuracy better than 3 mm within the shortest possible measurement time due to the limitation of railway skylight time. Among the existing Global Navigation Satellite System (GNSS) high-precision positioning technologies, Real-Time Kinematic (RTK) can achieve centimeter-level accuracy in a single epoch, while 3 mm positioning accuracy by static baseline processing requires at least one hour of observation time. This paper presents a GNSS Multi-Antenna Fast Millimeter-level Positioning (MA-FMP) method, which uses multi-antenna, multi-base and multi-epoch observations to achieve the effect of improving positioning accuracy and shortening observation time simultaneously. The innovation of the proposed method is using integration of RTK, multi-baseline adjustment, and Inertial Measurement Unit (IMU) attitude angle calibration to achieve a 3 mm positioning accuracy within a few minutes of static observation time. First, multi-antenna, multi-base and multi-epoch measurements are used to perform multiple static single-baseline RTK solutions. Then, multi-baseline adjustment including baseline vector geometric closure constraint and coordinate adjustment is used to improve the positioning accuracy. Finally, high-precision IMU attitude angles are used to calibrate the weighted average of multi-antenna result. A comprehensive experiment was conducted with real-world data on the Lunan high-speed railway in China, and results show that the proposed MA-FMP method using 3 minutes of static observation time and a distance up to 5 km from the base station can achieve a positioning accuracy of (1.5 mm + 0.3 ppm) and (3.0 mm + 0.5 ppm) in horizontal and vertical components, respectively. The MA-FMP method enables horizontal positioning accuracy of better than 3 mm within 3 minutes of static observation time, which can meet the requirements of rail track deformation monitoring within skylight of high-speed railway.