Strapdown inertial navigation system/celestial navigation system (SINS/CNS) is commonly used to high-precision missile navigation. Conventional SINS/CNS based on attitude measurement estimates the gyro drifts effectively. However, it does nothing about the accelerometer bias, which affect the positioning accuracy. Aiming at this problem, an improved SINS/Refraction and Kinematic-CNS (SINS/RK-CNS) method is proposed based on the indirect horizon reference measurement of star refraction. In this method, the relationship between refraction apparent height and missile position is deduced, and a new nonlinear integrated navigation model is established with the kinematic constraint of missile. Furthermore, corresponding observation noise model is developed to improve the positioning accuracy and the system robustness. Finally, simulations are conducted to test the performance with ballistic missile. Systematic position errors are calculated by traditional SINS/CNS method and SINS/RK-CNS based on EKF and UKF under different initial misalignment errors. The results confirm that the proposed SINS/RK-CNS has a significant effect on estimating the accelerometer bias. Compared to the traditional approach, three-axis position errors are decreased by 84.27%, 89.53% and 85.02%, respectively.
原文链接
分享
引用
