Abstract:
Objectives Aiming at the positioning requirements of vast users in the future, the correlation between the reference stations can't be removed when the traditional virtual reference station (VRS) observations are generated, and the interruption of individual reference station data will render VRS unable to be continuously generated.
Methods The undifferenced VRS algorithm addresses the correlation issue of correction values between reference stations and satellites. The undifferenced correction values at every reference station are mutually independent. When generating VRS corrections, an optimal subnetwork of reference stations can be selected based on real-time requirements, thereby eliminating the re-convergence of carrier-phase ambiguities for users triggered by data outages at individual reference stations. With fixed double-difference integer ambiguities, we assign reference integer ambiguities for all network reference stations and the reference satellite to achieve network-wide consistent undifferenced integer ambiguity estimation, and to generate the corresponding values.
Results Experimental verification is carried out for the positioning performance of the undifferenced VRS method. Four reference stations are used to generate VRS observations and construct two triangular subnetworks to resolve undifferenced integer ambiguities, with the grid spacing of VRS set to 10 km. VRS R1 is selected for algorithm testing, where user stations U1, U2 and U3 are arranged nearby. Differential positioning is performed using synthetic observations of R1, and user positioning accuracy is analyzed when a single reference station experiences a data outage.
Conclusions The experimental results indicate that the proposed undifferenced grid VRS method can maintain continuous VRS atmospheric corrections during reference subnetwork switching. In double-difference, the proposed method can also ensure continuous positioning results and realize centimeter-level high-precision positioning.
下载:
