Semi-physical Simulation And Nonlinear Optimizing Technology Of GNSS/INS Tightly Integrated Navigation System

GNSS(Global Navigation Satellite System)and INS(Inertial Navigation System)complement each other and become most widely used integrated navigation systems.However,nowadays the performance test of GNSS/INS integrated navigation algorithm and system are mainly realized through actual running testing and mathematical simulation,whose reliability and testing cost cannot be considered.Semi-physical simulation technology was applied to integrated navigation system,which fully combines the advantages of physical simulation and numerical simulation with high authenticity,reliability and universality.It makes very important significance to the research and development of military and civilian navigation system.In this paper,which has been subsidized by a project of Sichuan academy of aerospace technology,the theories and algorithms of GNSS/INS compact integrated navigation data fusion were regarded as the research objective,the integrated navigation system testing theories and methods were regarded as the research basis,the scheme design and implementation of the integrated navigation system based on hardware in the loop have been carried out.Initial alignment and calibration testing of triaxial turntable has been carried out,thus a semi-physical simulation platform based on triaxial turntable and GNSS satellite signal simulator have been built.Newton iteration moving horizon estimation algorithm has been proposed to eliminate interface caused by nonlinear factor move effectively,and effectiveness of the new algorithm has been verified through the semi-physical simulation system.The main works of the paper is summarized as follow:1.The principle and structure of different GNSS/INS combination modes were compared and analyzed.The navigation system models of tight combination modes were mainly studied,and their state equation and measurement equation were derived.Thus SINS/GPS tight integration simulatin software have been built by MATLAB.2.The error model parameters to be measured have been determined by learning the error model of inertial devices.According to the military standard and the specified application scenarios,the corresponding test outline was conducted,and the test parameters and methods were determined.The calibration of the gyro and accelerometer of the fiber-optic IMU(Inertial Measurement Unit)have been completed.3.In view of the defects in testing methods of GNSS/INS integrated navigation system at the present stage,the design and inplement scheme of integrated navigation system based on hardware in the loop simulation have been fulfilled,and a set of integrated navigation semi-physical simulation testing system have been preliminarily builded,which can be used for algorithm performance testing,device debugging,demonstration and verification of integrated navigation system.4.A moving horizon estimation algorithm in tightly integrated navigation system based on Newton iteration has been proposed,which can effectively reduce the adverse effects caused by the nonlinear characteristics of the navigation system.The performance requirements of inertial navigation devices are reduced,thus this algorithm has a broad application prospect in the popular application fields,such as mobile phone,vehicle navigation etc.5.The actual data were collected through on-board experiments and verified by the MATLAB based SINS/GPS tightly integrated navigation simulation software.Compared with the extended kalman filter,the new algorithm effectively improved the navigation and positioning accuracy and speed accuracy of the vehicle.6.Based on the built semi-physical simulation test system,a verification testing has been done.Compared with the unscented kalman filter,the new algorithm effectively improves the attitude,speed and position accuracy of the tight integrated navigation system in the dynamic scene.