Research On Vector-based GNSS/INS Deep Integration Navigation Technology

With the development of satellite navigation technology,its applications are becoming more and more widespread,but the complex application environment and demands bring great challenges to the traditional scalar receivers.In order to improve the performance of receivers,vector tracking technology is proposed.After years of development,numerous vecto r tracking methods have been studied in depth and their advantages have been recognized.However,due to the complexity of system implementation,it is still difficult to get practical applications.To solve the engineering application problems,th is dissertation was based on vector tracking theory and carrie d out research on the implementation methods of vector tracking loop(VTL)and vector-based GNSS/INS deep integration.The main research work is as follows.The dissertation firstly addressed the problem that the vector tracking theoretical framework is difficult to implement in hardware receivers,and proposed a vector tracking implementation framework based on the partial open-loop NCO(numerically controlled oscillator)control mode.During the navigation filter update interval,the code and carrier NCOs operate d in openloop mode;at the navigation filter update,the NCOs of all channels were corrected using the navigation state.Through this framework,the VTL could be realized by minor modification of existing hardware receiver.The implementation methods of each sub-module in the VTL were given,including pseudorange measurement,carrier frequency estimator,navigation update model,and navigation filter model.To improve the robustness of the VTL,a signal quality control method combining signal layer and information layer was proposed.In the signal layer,the carrier-to-noise ratio was used to eliminate weak signal satellites.In the information layer,robust filtering was used to suppress the interference of large error measurements to the navigation filter.In addition,a control strategy of loss-of-lock channels was given.By degrading the carrier tracking loop of the loss-of-lock channel to 1st-order phase lock loop,a fast reacquisition of the loss-of-lock signal was achieved.Based on the implementation framework of vector tracking,the dissertation further investigated the vector-based GNSS/INS deep integration method.The IMU(inertial measurement unit)was used to achieve more accurate motion state estimation,improve navigation accuracy,and c ould output attitude information.Meanwhile,the acceleration output from IMU was introduced into the carrier frequency estimator to counteract dynamic stress.The tracking capability of high-dynamic signals was improved.Three problems in the implementation of the navigation filter were considered,including lever arm effect compensation,carrier phase wind-up effect compensation,and data synchronization between GNSS and INS.The corresponding solutions were proposed.Because the acceleration auxiliary information contains a variety of errors,the influences of these errors on carrier tracking were investigated in the dissertation,including accelerometer bias and noise,attitude error,gyroscope noise,and environmental vibration.To address the problem of multi-system joint vector tracking,the dissertation designed a GPS+BDS joint VTL and verified the feasibility of joint vector tracking of BPSK(binary phase shift keying)and BOC(binary offset carrier)signals.For the problem of poor heading accuracy of the integrated navigation based on MEMS IMU(Micro-Electro-Mechanical System IMU),a dual-antenna vector-based deep integration method was proposed.By constructing the carrier phase difference measurement equation,the attitude error of inertial navigation could be corrected by using carrier phase difference directly.Not only could the calculation process of attitude angles be eliminated,but also the integer ambiguity of carrier phase difference be detected and fixed by using inertial navigation attitude information.At the end of the dissertation,the proposed methods were verified by three groups of experiments under actual signals.Firstly,the VTL was embedded in the hardware receiver based on FPGA+DSP,and the receiver was tested by GPS signal simulator.The test results were consistent with the simulation results.The algorithms met the real-time requirement.Secondly,a vector-based deep integration vehicle navigation experiment was carried out by using software receiver.The test results showed that the proposed vector-based deep integration method had excellent navigation performance in complex urban environment s.Thirdly,the influences of lever arm effect and carrier phase wind-up effect compensation were verified by an artificial motion with large angular velocity.Overall,the dissertation presented a more comprehensive study of vector tracking and vector-based deep integration methods,fully considering the technical details in the implementation process.The proposed methods were validated by both simulation and experiment,which ha ve strong values of theoretical reference and engineering application.