Research On Key Technologies Of Receiver Tracking Algorithm Under Multi-Constellation Multi-Frequency GNSS/INS Deep-Coupled Architecture

With the development and modernization of the global navigation satellite system(GNSS),GNSS features multiple constellations,multiple signal frequencies,and multiple signal modulation types to provide users with more and more abundant measurement information,flexible processing methods and better positioning,navigation,and timing service.GNSS/INS deep-coupled navigation technology deeply integrates receiver baseband and inertial navigation system(INS)information,which is an important method to improve the performance of GNSS receivers.However,most of the existing GNSS/INS deep-coupled receiver tracking technology is researched based on traditional signals,simplified algorithm models and simulation scenarios.In the actual urban environment,they still face problems of weak signals,signal occlusion and blocking,and insufficient number of visible satellites,and complex practical application problems such as multipath/NLOS signal interference.In order to fully tap the diversity of the receiver's internal channels,the diversity of modern GNSS multi-frequency observation signals,and the integration of INS-assisted information,and further enhance the receiver's tracking and navigation positioning performance,this paper takes the environment of urban canyon occlusion and other complex scenarios as a scenario to develop multi-constellation and multi-frequency GNSS/INS deep-coupled receiver tracking key technology,and carry out effectiveness verification work through the hardware simulator and actual vehicle experiments.The main work of the paper is as follows:(1)Research on vector joint tracking algorithm based on GNSS receiver channels' diversity.The system mathematical model characteristics of both typical coherent and non-coherent vector tracking loops are analyzed firstly.Then,the receiver tracking channel diversity and its inherent channel coupling mechanism are deeply analyzed on the basis of the equivalent noise bandwidth of the vector tracking loop.For coherent and non-coherent vector tracking having different advantages in accuracy and robustness,an adaptive hybrid pre-filtering vector tracking loop(AHVTL)is proposed,A detailed adaptive hybrid pre-filtering scheme design is given to achieve mutual assistance of receiver tracking channels.For the defect of poor tracking robustness of vector tracking,a robust processing filtering algorithm based on robust adaptive factor is designed to avoid receiver tracking channels' crossing "pollution",which can effectively improve receiver tracking accuracy and robustness.(2)Research on multi-frequency joint tracking algorithm based on GNSS signal's diversity.The characteristics of GNSS signal diversity and the new characteristics of modern signals are fully explored and analyzed.Then,research on the reception processing mode of GNSS modern signals is carried out,the theoretical Cramér-Rao lower bound of the spreading code tracking error is given,and the two main reception types i.e.,matched reception and mismatched reception processing mode and its tracking error are analyzed.Taking B1 C and B2 a new modernized signals of Bei Dou navigation satellite system: phase 3(BDS-3)as an example,the design based on the optimal linear combination Doppler-aided tracking loop(OLC-DATL)and Kalman Filter-based Joint Tracking Loop(KF-JTL)are researched.For coherent integration time of pilot and data joint tracking due to the problem of navigation bit limitation in data channel,the non-coherent carrier phase detection algorithm is introduced to realize the expansion of the pilot and data joint tracking loop integration time,and also the smooth sliding window design is used to increase the loop update rate and reduce the loop dynamic response delay.Pilot/data channel joint tracking and multi-frequency joint tracking are valided effectively to improve receiver's tracking sensitivity and anti-interference performance.(3)Research on GNSS/INS deep-coupled navigation and tracking algorithm based on improved hybrid pre-filtering processing.First,the basic theory of the GNSS/INS deep-coupled navigation system is given,and the architecture and mathematical model of the centralized and cascaded deep-coupled navigation system are analyzed and compared.The problem of the pre-filter model of the existing cascaded deep-coupled architecture is analyzed,A fifth-order cubature Kalman filter(5th-CKF)algorithm is proposed to solve the highly nonlinear problem of the coherent pre-filter model;Also,aiming at the influence of the RF front-end and filter bandwidth on the smoothing of the code autocorrelation curve,a sixth-order polynomial fitting code autocorrelation curve is designed to better characterize the actual code autocorrelation characteristics.For the non-coherent pre-filter model,in order to reduce the code and carrier noise correlation terms generated by the discriminator function,the code and carrier state and measurement are decorrelated and optimized to the greatest extent,and the measurement noise variance is adjusted in real time.In addition,by simplifying and separating coherent and non-coherent pre-filter models,the signal amplitude state item in the state variables of the system reduces the dimension of the system state vector,and at the same time avoids the initial setting of the signal amplitude of the filter and speeds up the convergence rate of the filter.Finally,based on the improved coherent and non-coherent pre-filtering models,an adaptive GNSS/INS hybrid deep-coupled navigation(Hybrid-DC)algorithm is proposed and detailed algorithm models and parameters are also given.The proposed Hybrid-DC algorithm obtains the best tracking and navigation performance by fusing the advantages of coherent and non-coherent modes: i.e.,the robustness of non-coherent mode makes the Hybrid-DC algorithm track more satellites,especially in an environment where the GNSS signals are severely blocked,while the coherent mode helps the Hybrid-DC algorithm improve its tracking and positioning accuracy.(4)Research on multi-constellation multi-frequency GNSS/INS deep-coupled information fusion and multipath mitigation algorithm.On the basis of analyzing the effects of multipath/non-line-of-sight(NLOS)signals on the receiver and the analysis of multipath mitigation and interoperability performance of modern GNSS signals,Research on multi-constellation multi-frequency GNSS/INS deep-coupled information fusion algorithm is carried out,the details of baseband architecture,pre-filter model and navigation filter information fusion algorithm are designed.In the signal domain,L1 band(L1C,B1 C,E1)and L5 band(L5,B2 a,E5a)interoperability modernization modern signals are selected,and the ability of high code rate and binary offset carrier modulation(BOC)signals to suppress multipath are simulated.Further mitigation of multipath signals is achieved through the design of narrow correlator spacing in the baseband domain,the NLOS signal detection and rejection algorithm of deep-coupled system,and the expansion of the coherent integration time in the Doppler domain,so as to improve the system's availability,continuity,navigation accuracy and robustness,and also the capabilities of multipath mitigation and NLOS signal detection and rejection.