Research On Shipborne IMU/GNSS Tightly Integrated Navigation Method Based On MEMS

With the rise of the concept of unmanned ships and the development of related technologies,higher requirements are put forward for ship positioning,navigation,and timing(PNT)information.Global Navigation Satellite System(GNSS)has the advantages of all-weather and no-error accumulation and is the main source of PNT information in the maritime field at present.However,due to the influence of shielding,multipath,intentional interference,and deception,the performance of the system has the risk of serious decline.It has become a key scientific problem to effectively deal with GNSS vulnerability and ensure the accuracy and continuity of shipboard PNT data.In this thesis,the combined structure and data fusion method of Inertial Measurement Unit(IMU)and GNSS based on Micro Electro Mechanical System(MEMS)technology is studied.The research results can effectively meet the development needs of unmanned ships in the future maritime field..In this thesis,the state model and observation model based on pseudo-range and pseudo-range rate are deeply studied,and the nonlinear mathematical model of error state under MEMS-IMU/GNSS compact combination structure is constructed.Aiming at the characteristics of poor stability and fast drift of MEMS,a closed-loop integrated system based on feedback complementary structure was designed.The system state and output were corrected while the system error was estimated,which effectively restrained the rapid divergence of MEMS.Then,the simulation of the tight combination system based on the Extended Kalman Filter(EKF)algorithm is carried out,and it is proved that the established mathematical model and combination structure can not only improve the positioning accuracy but also provide the attitude angle information with the reference value.In addition,in order to solve the error model imprecise problem caused by the linearization method in EKF that preserves the Taylor series expansion term to the first or second order.In this thesis,a Symmetric Square Root Unscented Kalman Filter(SS-SRUKF)algorithm is proposed.The Square Root of the covariance matrix is transmitted by QR decomposition and Cholesky decomposition.The computational efficiency and numerical stability are improved.Then,based on the error characteristics of tactical and consumer MEMS,the SS-SRUKF algorithm is simulated with 9 and 3 visible satellites,respectively.The results show that the proposed algorithm can effectively improve the accuracy of the navigation solution,and provide the robustness of a short-time bridge when the number of visible satellites is not enough to provide a GNSS navigation solution.Finally,in order to verify the applicability of the aforementioned theoretical research algorithm,a Java-based MEMS-IMU/GNSS shipside navigation system test platform is developed to provide application examples for the future intelligent all-source navigation system of unmanned ships.Based on the analysis of the overall demand of the platform,the detailed design scheme and processing flow of the space-time unified module,the fusion and evaluation module,and the interface display module are given.The ship motion data generated by the simulator under third-order sea conditions are used to verify the rationality of the design scheme of the test platform and the correctness of the function.Based on the filtering uncertainty data,the EKF and SS-SRUKF filtering algorithms are analyzed and evaluated,which further proves the advantages of the SS-SRUKF algorithm in accuracy and stability.