Information Processing Of SINS/GPS Integrated Navigation System Based On DSP

With the development of modern control theory, the requirement of navigation accuracy is forward to higher. Strap-down Inertial Navigation System(SINS) and Global Positioning System(GPS) are two kinds of navigation systems which are often used at present. SINS has a strong autonomy, is not vulnerable to interference, but the error of the navigation solution can be accumulated over time. GPS positioning accuracy is high, more stable, but it is easily disturbed by large obstacles. According to the complementary advantages and disadvantages of SINS and GPS, this paper studies the SINS/GPS integrated navigation system based on DSP.Firstly, the working principle of SINS and GPS are analyzed, the two sources of error are summarized. These provide a theoretical basis for the design of integrated navigation system. By comparing the loose combination mode and tight combination mode, the whole structure of the system is designed by tight combination mode. Then the integrated navigation system is designed, the design of hardware system includes data acquisition module, data processing module and peripheral circuit of DSP. The design of the system software part includes the choice of the software platform and the design of the whole process of the software.Secondly, in order to optimize data processing of integrated navigation system, improve the precision of navigation. This paper explains the SINS initial alignment method, navigation data of time synchronization and navigation information fusion technology in detail. Due to the state equation of SINS/GPS composite system has nonlinear form, the phenomenon of filtering divergence is existed in the Kalman filter(KF) method. Therefore, the cubature Kalman filter(CKF) method is proposed. Cubature Kalman filter using the cubature rule to calculate the Gauss weighted integral, can better suppress the nonlinear noise. The above two filtering algorithms are simulated in the MATLAB environment. The results show that the filtering effect of CKF is better than KF, and the estimation accuracy is also improved.