Improved Kalman Filter Algorithm of GPS.doc

Improved Kalman Filter Algorithm of GPS/INS Integrated Navigation without GPS Signal ZHU Lixin1, MENG Yan2 (1.Electronic Engineering Institute of PLA, Anhui 230037,China) (2.No.72671 Unit of PLA, Jinan 250022,China) Abstract: To solve the problems that the navigation accuracy of GPS/INS(Global Positioning System/Inertial Navigation System) tightly integrated navigation system is lower under GPS signal invalidation circumstances, and the model error and calculation error exists in the application of CKF(Cubature Kalman Filter), a strong tracking SRCKF(Square-Root CKF) is proposed under the background. By stressing the function of new data artificially, the algorithm improves the robustness when the model is uncertain, and the square-root strategy ensures the positive definiteness and symmetry of covariance matrix. Simulation results show that the improved algorithm can improve the navigation accuracy and the ability of adapting to the complex environment, and have better efficiency under GPS signal invalidation circumstances. Key words: GPS/INS tightly integrated navigation, CKF, strong tracking filter, square-root strategy, navigation accuracy 1. Introduction The navigation accuracy of GPS/INS integrated navigation system overcoming the respective defects is higher than the accuracy of the two systems working separately. When the receiver is under the great angle motion and the complex circumstances of the city and the valley, satellite signals are disturbed and shielded easily. On the situation, observations provided by GPS/INS tightly integration utilizing less than 4 satellites in the measurement of pseudorange and pseudorange rate leads to integrated model working well. Considering the nonlinear characteristic of GPS/INS tightly integrated navigation model, a hybrid linear/nonlinear model based on linear state equation and nonlinear measurement equation is proposed. EKF(Extend Kalman filter) and UKF(Unscented Kalman filter) are classical nonlinear filters. EKF