GNSS/INS Tightly-coupled Algorithm And Performance Analysis In Multiple Signal-blocked Environment

Vehicular mobile surveying technology has become one of the cutting-edge technologies in the discipline of surveying and mapping,and the core is positioning and attitude determination technology.The combination of Global Navigation Satellite System(GNSS)and Inertial Navigation System(INS)can provide more accurate navigation information and become one of the best solutions to solve the problem of positioning and attitude determination of a vehicular mobile surveying system.Focusing on the multiple GNSS signal-blocked environment,the paper focuses on the navigation system of the differential relative positioning of GPS/BDS/GLONASS and the INS tightly-coupled.The main research work is as follows:(1)The GPS,BDS,and GLONASS observations are combined to study the relative techniques of dynamic carrier phase differential relative positioning.Using the measured data,the results show that when the cut-off elevation angle is less than 35°,the fixed rate of single-epoch ambiguity is about 97%and the fixed error rate is within 0.12%.With reference to Inertial Explorer 8.60,a high-precision differential tightly coupled post-processing software,the horizontal direction accuracy is better than 0.02 m and and the vertical direction accuracy is better than 0.05 m.(2)The INS differential equation and error equation,the dynamic model and observation model of GNSS/INS tightly-coupled are deduced strictly.The idea of decentralized filter is introduced and the ambiguity is fixed separately,so the simplest extended Kalman filter(EKF)is constructed,and RTS smoother algorithm is studied.The above algorithm was validated by using several sets of measured data.The position accuracy of EKF forward filter is about 0.02 m,the velocity accuracy is about 0.01 m/s,the accuracy of the roll and pitch angles is about 0.007°,and the heading angle accuracy is about 0.07°.Compared to the loosely-coupled,the tightly-coupled position,speed,and attitude accuracy are improved.After RTS smoothing,the position accuracy is improved from 9%to 55%,the accuracy of the roll and pitch angles is improved from 10%to 30%,and the accuracy of the heading angle is improved from 20%to 60%.(3)The methods of INS-assisted cycle-slips detection and repair and INS-aided ambiguity fixing are studied.According to different navigation stages,a variety of ambiguity fixing strategies are used.Using INS information to replace pseudoranges in the combined observations of Melbourne-Wiibbena,pseudorange noise and multipath effects are eliminated.The experiment simulates small cycle-slips,large cycle-slips,and dense cycle-slips respectively,all of which can be detected and successfully repaired.Within 150 s of the simulated GNSS signal interruption,the corresponding INS calculated position error is within 3.5 m,and the cycle-slips detection and repair algorithm is still valid.