Research And Application Of Consumergrade PPP/INS Integration Technology

With the upgrading of the Global Navigation Satellite System(GNSS)in the past decade,high-precision navigation and positioning has gradually expanded from traditional fields such as military and surveying to new consumer market areas such as wisdom logistics and vehicle-borne navigation.This is not only due to the breakthrough of navigation hardware in the direction of low cost,small size,low power consumption,and high precision.At the same time,the rapid development of multi-sensor fusion algorithms has also laid a solid theoretical foundation for consumer-grade highprecision navigation and positioning applications.On the one hand,GNSS Precise Point Positioning(PPP)technology integrates the technical advantages of Standard Point Positioning(SPP)and relative positioning.It is capable of achieving decimeter to sub-meter level accuracy in benign GNSS conditions using consumer-grade single-frequency(SF)GNSS receivers;on the other hand,the Inertial Navigation System(INS)with Micro-Electro-Mechanical System(MEMS)Inertial Measurement Unit(IMU)can provide continuous and stable positioning information in the case of frequent GNSS signal degradation and blockage.Therefore,the consumer-grade PPP/INS integration,or SF-PPP/MEMS-INS,has great potential in the civilian field of high-precision positioning.However,many issues need to be addressed in the process of promoting the consumer-grade PPP/INS integration to the market,such as the observation gross error detection and the high noise rejection,the accucary uncertainty of consumer-grade PPP in complex areas,and the coupling issue of the internal parameter from the consumer-grade PPP/INS integration.This paper aims to establish a set of real-time consumer-grade PPP/INS integration with both software and hardware to meet the stable decimeter(open-sky)to sub-meter(short-term blockage in sub-urban areas)to meter level(dense blockage in the downtown)positioning accuracy.The main contributions are as follows:(1)The simple and practical methods in pre-processing of single-frequency GNSS observations are proposed.Code gross errors are detected by using doppler observations and post residuals in SPP simultaneously,and phase cycle slips are detected by using code and doppler observations simultaneously.In addition,the specific detection thresholds for GNSS observations from consumer-grade receivers are directly given,which provides the convenience for interested users.(2)The consumer-grade multi-GNSS PPP system in the user end is established,and the positioning performance of PPP in different areas is discussed and analyzed.Both GPS/GALILEO based and GPS/GLONASS/GALILEO based PPP can provide stable decimeter to sub-meter level positioning accuracy in both horizontal and vertical directions in open-sky and sub-urban areas.In the downtown,it can still provide the horizontal positioning accuracy of 10.0 m.(3)In order to solve the systematic positioning error in GLONASS-only based PPP the GLONASS Inter-frequency Code Bias(IFCB)is analyzed.We propose to use University of Calgary(Uof C)model for GLONASS IFCB estimation in SF-PPP,then verify the advantages of Uof C model over the traditional single-frequency PPP model.Further,the stability of GLONASS IFCB in the low-cost SF receiver u-blox M8 T is analyzed and investigated.Finally,both static and kinematic SF-PPP accuracy in GLONASS-only have significant improvements,reaching the same level as GPS-only,after estimating or calibrating GLONASS IFCB.(4)The local ionospheric delay model based on second-order polynomials is proposed to improve the accuracy and the calculating efficiency of the ionospheric delay in the user end,further improve the SF-PPP accuracy.The positioning accuracy remains decimeter-level steadily in a local coverage,no matter whether the ionospherc condition is quiet or active.(5)The method is proposed to address the issue of the short-term interruption of realtime orbit and clock corrections.The method generates the orbit and clock by directly extrapolating based on the orbit and clock corrections from the first epoch and the broadcast ephemeris.The accuracy of the orbit and clock are maintained within 50 cm and 0.55 ns within one hour,respectively.In the meanwhile,the accuracy of the kinematic SF-PPP can still be maintained at sub-meter level.(6)The consumer-grade PPP/INS integration system is established to solve the issue of low accuracy and the inability to locate in PPP when encountering severe multipath effects and GNSS outages.In complex urban areas,the consumer-grade PPP/INS integration system can stably provide positioning accuracy from decimeter to meter level,speed accuracy of several decimeters per second,and attitude accuracy of about0.5°.In the meanwhile,the horizontal positioning accuracy in open-sky areas in 3minutes is 0.433 m,and the maximum positioning error does not exceed 0.749m;whereas 0.570 m and 1.403 m for continuous bridge blockages in sub-urban areas in 7minutes;2.766 m and 5.244 m for the tunnel under the bridge in 1 minute;3.006 m and7.411 m for the downtown with dense blockages in 8 minutes;3.265 m and 5.936 m in underground parking lot in 4.5 minutes.