Research On Multi-GNSS/INS Tightly Coupled Positioning Method Considering Inter-system Bias

The arrival of the intelligent era requires navigation and positioning technology with higher accuracy and reliability.Aiming at the problem of unreliable system model and low accuracy of the GNSS/INS integrated navigation system in the current complex urban environment,the paper studies the multi-GNSS/INS tightly coupled mode,which takes into account the inter-systems bias.Based on the multi-GNSS system tightly coupled positioning model,the INS is combined to establish a more reliable system model and improve the accuracy and stability of the GNSS/INS tightly coupled combination system.This paper researched on three aspects: firstly,multi-GNSS system positioning model construction;secondly,GNSS cycle slip detection problem;finally,multi-GNSS/INS tightly coupled filtering algorithm,etc.The main work is as follows:(1)The paper derived a multi-GNSS unified positioning model which takes into account the inter-system bias for multi-GNSS positioning.Based on the original single-system navigation observation model,the combined observation model between different GNSS systems was expanded.In addition,the multi-system combined data was considered by the difference in observation values.Consequently,the algorithms makes up for the low accuracy and insufficient availability of the single GNSS system and multi-GNSS loosely coupled mode that were limited by the harsh environment during positioning.(2)Aiming at the problem of frequent cycle slips of GNSS carrier phase observations in a dynamic environment,which reduced the accuracy of the observations.The paper has verified the INS-assisted single-difference carrier phase cycle slip detection method based on robust estimation through experiments,and proved that the method can effectively detect the occurrence cycle and calculate its size.(3)The theoretical model of the multi-GNSS/INS integrated navigation system in a complex environment has difficulties in matching the actual model,resulting in a decrease in filtering accuracy or even divergence.The paper studied two improved nonlinear filtering algorithms,used on-board experiment to verify.Validating by experimental analysis that the square root CKF algorithm hasd a certain increase in accuracy compared with the standard EKF algorithm,but there was still a problem of large deviation in the solution results;the time-varying noise estimation STCKF algorithm could not only improve the robustness of conventional nonlinear filters to uncertain system models and the ability to track sudden changes in the system state,but also effectively suppressed the influence of abnormal observation information on the navigation solution,and improved the accuracy and stability of the navigation system.