Key Algorithms Of The Information Processing In The Multi-constellation Integrated Navigation

Global Navigation Satellite System (GNSS) is the indispensable nationalinfrastructure for supporting the national economic construction and maintaining themilitary security, and becomes competing hotspot of today’s major world powers. Atpresent, there are three global satellite navigation systems providing navigation andpositioning services for the users, Chinese Compass satellite navigation system,American GPS and Russian GLONASS, and the European Galileo system is alsostepping up construction. However, even the most sophisticated satellite navigationsystem-GPS, some defects of the system have emerged as showing many advantages atthe same time. For instance, GPS can not meet the requirements as a single navigationsystem for navigation of the objects in high dynamic, due to the limitation of systemaccuracy and reliability. The multi-constellation integrated navigation system can usethe observations come from multiple systems at the same time, so it has more availablesatellites, better coverage, higher positioning accuracy, and higher reliability, and isgradually becoming one of the hot research field of satellite navigation.This paper focuses on two aspects of multi-constellation integrated navigationinformation processing: the available satellite selection and the filtering basednavigation solution.Firstly, multi-constellation integrated navigation satellite selection algorithm isintroduced. Due to the large difference of measurement accuracy among satellites, andthe large number of selectable satellites, the traditional satellite selection algorithm is nolonger applicable for the multi-constellation integrated navigation satellite selection.Regarding to this problem, a multi-constellation integrated navigation satellite selectionalgorithm based on Weighted Position Dilution of Precision (WPDOP) contribution isproposed. Compared to the traditional satellite selection algorithm, the algorithm hasless amount of calculation and higher positioning accuracy, and it is engineeringpractical. At the same time, derived through mathematical formulae, an optimizedalgorithm for WPDOP and WPDOP contribution amount in the multi-constellationintegrated navigation system is proposed, which greatly reduces the computationalcomplexity and has a certain application value.Secondly the multi-constellation integrated navigation solution algorithm isintroduced. When the multi-constellation integrated navigation system is applied to thehigh-dynamic environment, with the influence of the acceleration changing in a widerange and the properties for kinematics noise and measurement noise difficult to beascertained exactly, the estimation precision of the general federated filtering decreasesseriously, and the filtering even became divergence. Regarding to this problem, thispaper applies strong tracking filter to fault-tolerant federated filter, forming a fault-tolerant federated strong tracking filter (FTFSTF). The algorithm proposed isapplied on Compass/GPS/GLONASS multi-constellation integrated navigation system.The simulation results show that the algorithm is flexible, strong fault tolerant, andbetter to track high-dynamic targets, and has significantly improved the precision andreliability of the multi-constellation integrated navigation system. Simultaneously,owing to the combined application of the no-reset mode federated filter and somepractical algorithms including the one-step suboptimal calculation method for fadingmatrix and the residual Chi-square detecting method, the FTFSTF algorithm hasmoderate calculation complexity, and it is easily implemented and valuable forengineering application.The results of this research can provide valuable references for themulti-constellation integrated navigation information processing engineering.