Research On The Key Technologies Of Multi-GNSS Integrated Navigation And Location

With the modernization of the Global Navigation Satellite Systems(GNSSs),the techniques of multi-frequency multi-system navigation and positioning will become the future development trend of satellite navigation.First,combining observations from various satellite constellations significantly increases the number of available satellites and the achievable accuracy,reliability and availability,especially in constrained environments.Second,the combination of multi-frequency observations can be used to eliminate or mitigate individual error sources(ionosphere,troposphere/orbital,and noise/multipath),increase reliability of carrier ambiguity resolution and improve cycle slip detection and repair.In order to integrate the above advantages and realize multi-frequency and multi-system integrated navigation and positioning,this paper studies the key technologies of multi-GNSS integrated navigation and positioning technology,including multi-GNSS common characteristics and performance analysis,multiple constellation integrated navigation system,Multi-GNSS stochastic model refinement,Multi-GNSS loose,tightly integrated positioning technology,inter-system deviation influence analysis,mixed double-difference phase observation linear combination model and characteristic analysis,real-time precision relative positioning.The main achievements and innovations are as follows:1.Based on the study of the common characteristics of the satellite navigation systems,the similarities and differences between the four navigation systems,the standardization of space-time datums among the systems,and the performance of Multi-GNSS are studied respectively.For quantitatively researching the performance improvement of different levels of GNSS Multi-constellation Integrated Navigation system,the paper constructs the corresponding quantitative indicators to analyze the performance improvements of the system quantitatively.Through the experiment of simulation,it verifies the validity of indexes,and analysis the rule of the performance improvements of the system along with combination levels and cut-off elevation angle change,and proposes the optimal system combination at all levels.It demonstrates that the high-level of multi-constellation system not only improves the flaws of single and low-level configuration,also largely enhances the performance of constellation.2.Based on the analysis of relationship between the satellite elevation as well as the geometry dilution of precision(GDOP)and User-Equivalent Range Error(UERE),a new fast satellite selection algorithm applicable to GNSS receiver was presented with comprehensive considering satellite space geometric distribution and observation error on the result of positioning.In this algorithm,first,the simplified positioning precision evaluation function of elevation and the selecting satellites number model are built with the consideration of GDOP and UERE,through which the minimum elevation thresholds and the number of selecting satellites were quickly decided according to the precision of user's requirement.Then,the partition-weighted method is used to select all the satellites that are greater than the minimum elevation threshold.Parts of satellites are excluded through sorting,differentiating to fulfill satellite selection.Experiment results show that the algorithm has much lower complexity compared to the WPDOP algorithm,and the selection effect is better than the optimal PDOP algorithm.3.By introducing the information fusion theory,the multi-GNSS data fusion strategy is analyzed systematically,and the multi-GNSS fusion weight ratio problem is studied.Analysis of Multi-GNSS Single point positioning stochastic models.As to the accuracy of stochastic model in combined positioning,a method combining priori weighting based on UERE and posterior weighting based on variance estimation is proposed.Experimental results showed that: Compared with single-system,Multi-GNSS have more available satellite in single epoch and smaller GDOP,which makes the positioning accuracy,integrity,availability and other navigation performance has greatly improved.Compared with prior weighting or posteriori variance estimation weighting,the weighting method can effectively improve the accuracy of Multi-GNSS positioning and reduce the number of iterations.4.Multi-system double difference combination methods have been studied in detail,including loose integration(each of the systems uses their own independent pivot satellites and none of the double difference observations will be formed across the two systems)and tight integration(the two systems use the same pivot satellite and support creating double differences between satellites of different systems).By transforming the difference strategy,we unify the multi-GNSS combined double-difference observation model,and then design the corresponding model for the loose integration and the tight integration double differential relative positioning respectively.Finally,the effectiveness of the difference strategy and model is verified by experiments.5.The double-differential relative positioning of the multi-GNSS loose combination with short and long baselines was studied.The experimental results show that the availability and ambiguity fixed success rate and accuracy of the Multi-GNSS loose combination relative positioning is very similar to that of single-system GPS or BDS,but it is better than single-system GLONASS.But the combination of positioning results are not the same with the higher the quality of the better,but also consider the introduction of data quality or reasonable weighting.High-cut height angle environment,the single system can not achieve the whole day positioning,and multi-system combination can maintain the advantages of more visible satellite number,can be more complex environment to provide a stable navigation and positioning services,and has a high Ambiguity fixed success rate and accuracy.Under the long baseline,the combined system can maintain high positioning availability,but the ambiguity fixed success rate increases and then decreases with the increase of the cutoff height,which leads to the new idea of eliminating the influence of medium and long baseline relative to eliminate the atmospheric delay.6.The tight integration has a higher redundancy than the loose integration,but introduces an inter-system bias(ISB).The stability and processing strategy of ISB are analyzed and the relative positioning performance of two integrations is analyzed and compared.The results show that the ISB has good time stability,which is related to the combined receiver and can be modeled by introducing the parameters.In the relative positioning of multi-GNSS co-frequency combinations,the influence of the same type of receiver pair,code and phase fraction ISB can be neglected.However,when different types of receivers are used to locate,the elimination of ISB can significantly improve the success of ambiguity rate.7.A tightly combined observational model assumes a single reference satellite for all observations from different GNSS systems.However,this model will no longer be used,in case of lack of coinciding frequencies and hence an inter frequency bias.We analyze the effect of different frequencies on the model of different-frequency inter-system mixed double differences(DFMDDs),and derive a minimum constraint of inter-frequency difference.To mitigate inter-frequency difference,we propose a new idea: using combination of phase observations from different GNSS systems to obtain combined virtual observations with the same or similar frequency.Subsequently,a new carrier phase linear combination method is proposed for the DFMDD,called limited lane number search method,which increases the minimum constraint of inter-frequency difference.We demonstrate the feasibility of the search method from a geometrical perspective.Finally,for the DFMDD,we provide some excellent combinations.The optimal independent integer combinations are suggested for short baselines and long baselines.