The Calibration Of Strapdown Fiber Optic Gyrocompass Using SINS/GPS Intergration

Strapdown system is the general trend of the development of INS (Inertial NavigationSystem); The fiber optic gyroscope as a new solid-state inertial instrument has a lot ofadvantages, representing a new derection of the development of inertial device; The navalvessel’s compass system can provide not only course information for ship navigation, butalso horizontal attitude information for radar, sonar and weapon control system, and it hasbeen highly valued by the navy of every country for many years. So it’s of vital importanceto research and develop high-accuracy strapdown fiber optic gyrocompass system. Thispaper mainly studies the combined calibration of strapdown fiber optic gyrocompass systemand data fushion methods using in the calibration.In this paper, the main content is as follows:1、The development of strapdown system, fiber optic gyroscope and strapdown fiberoptic gyrocompass home and abroad is overall summed up, as well as the integrated system.By analyzing the errors of strapdown fiber optic gyrocompass system, the combinedcalibration plan under undamped inertial working state researched in this paper is designed.Then, the error model of strapdown fiber optic gyrocompass system is established and theerror analysis of GPS is given. At last, the open-loop adjustment based on position andvelocity integration is choosen as the integrated adjustment way of combined calibrationsystem.2、Kalman filter’s key role in combined calibration is analyzed and traditonal linearKalman filter equations are given. Then, the mathematic model is particularly introduced,namely the state equations and measurement equations. Based on this model the simulationof combind calibration system using Kalman filter is desinged and the feasibility andeffectivity of the proposed combined calibration plan is proved. In general, the system isnonlinear, and the traditonal linear Kalman filter is not suitable to the nonlinear model. Soseveral typical nonlinear filter methods including EKF、UKF and particle filter areintroduced and then the simulation and analysis on the filtering accuracy of EKF and UKFis estblished, the simulation results showing that the filtering accuracy of UKF is higherthan EKF under the same condition.3、The forementioned nonlinear filtering methods including EKF、UKF and particlefilter are summed up. Then based on UKF algorithm the unscented particle filter (UPF)algorithm is introduced. Nextly, the innovation-based and residual-based adaptive kalman filtering are introduced. Then combined with UPF algorithm this paper studies the newnonlinear filter algorithm——adaptive unscented particle filter (AUPF) algorithm, and drawthe flow chart of AUPF. At last, a series of simulations are designed, including the systemperformance comparison between UKF and AUKF algorithms under benign and highlyreflective noisy signal environment, the system performance comparison between the AEKF,AUKF and AUPF algorithms under highly reflective noisy signal environment andestimation accuracy comparison of AUPF algorithm with different numbers of particles. Thesuperiority of AUPF compared with other nonlinear filtering methods is varified through thesimulations.