Research Of GPS/BDS2/BDS3 PPP/INS/Vehicle Auxiliary Information Tightly-Coupled Algorithm

With the development of emerging fields such as intelligent driving and mobile robots and the increasing demand for location services,high-precision,high-reliability,continuous and robust positioning technology has attracted the attention of academia and industry in recent years.The comprehensive networking of the third-generation Beidou system has promoted the further development of multi-mode and multifrequency PPP technology.However,in dynamic and complex environments such as urban canyons,bridges,and tunnels,as satellite signals are frequently out of lock,which makes PPP positioning frequently reconvergence,accuracy,reliability,and continuity of the system get worse.At present,no single positioning method can meet the needs of multiple scenes and multiple functions.What kind of sensor to choose and how to fuse various observation information together are worthy of in-depth study.This article has carried out research on GPS/BDS2/BDS3 PPP algorithm and PPP/INS/vehicle auxiliary information tightly-coupled algorithm,design and realize an algorithm framework that integrates GNSS PPP solution,GNSS/INS loose-coupled algorithm,and GNSS/INS/vehicle auxiliary information robust and tightly-coupled algorithm.Finally,the experiments are fully verified in the open scene and complex scene.The main research work and contributions of this paper are as follows:1.The GPS/BDS2/BDS3 PPP algorithm is derived based on the basic theory of GNSS and the ionosphere free combination model.The accuracy of the algorithm is verified by the on-board dynamic test,and the Beidou third-generation system and its new signals are tested and evaluated.The analysis believes that:(1)After the introduction of the BDS3 system,the number of satellites has increased,the geometric configuration has also been optimized,and the PPP positioning accuracy has been significantly improved;(2)Thanks to the advancement of the new signal design,the accuracy the BDS3 B1C/B2 a combined positioning result is about 20% higher than the B1I/B3 I combined positioning result,and the convergence speed is about 40% higher.2.Based on the basic theories of inertial navigation and integrated navigation,using sequential square root filtering,the PPP/INS tightly-coupled algorithm is derived and designed.In the open sky,the horizontal position accuracy of PPP/INS tightly-coupled algorithm can reach about 0.05 m,the elevation accuracy can reach about 0.25 m,and the yaw accuracy can reach 0.1°.However,in a complex environment,due to frequent loss of satellite signals or the influence of factors such as multipath,although the PPP/INS tightly-coupled navigation has better positioning continuity,the reliability of the positioning results is poor.On the basis of the PPP/INS tightly-coupled algorithm,the vehicle-mounted auxiliary information constraints are added.In the algorithm design,the INS/vehicle-mounted auxiliary information loose combination is solved first,and then the tight combination is calculated with the original GNSS observation information.On the one hand,it can suppress the rapid divergence of inertial navigation errors when the GNSS is completely out of lock,and on the other hand,it can speed up the reconvergence of the positioning results.This article is tested by simulating the loss of lock of GNSS signals.The experimental results show that:(1)During the GNSS interruption,non-holonomic constraint and odometer assistance can greatly suppress the divergence of positioning errors,and the accuracy can be increased by about 87%,Within 60 s,the positioning accuracy can maintain about 1.5m;(2)The strategy of fusing the vehicle-mounted auxiliary information with the INS firstly,and then combine with the original GNSS observations can speed up the PPP reconvergence speed by about 86%.3.Add robustness to the tightly-coupled algorithm.In sequential filtering,the chisquare test is performed by using the inspection quantity of the innovation structure,which can not only detect each observation under the same epoch individually,but also has a more accurate error detection capability.The test results in complex urban scenes show that the robustness algorithm further improves the reliability of the algorithm.In the case of serious interference of GNSS signals such as viaduct and soundproof shed,the results of positioning can keep smoothness and prevent the abnormal positioning fluctuation.