All-Parameter System-Level Calibration for SINS/CNS Based on Global Observability Analysis

Before using the strap-down inertial navigation system (SINS)/celestial navigation system (CNS) integrated navigation system, the calibration of its error parameters is a necessary process to improve the system accuracy. In this article, an all-parameter system-level calibration method has been proposed, which utilizes Kalman filtering for simultaneously estimating the bias, scale factor, misalignments of inertial measurement unit (IMU), and installation errors of the star sensor. The observation has been constructed using the inertial navigation output information and the starlight vector measured by the star sensor. The observability of the error parameters has been analyzed theoretically, and the optimal excitation conditions for each error parameter are obtained. The effects of the angle between the starlight vector and the optical axis of the star sensor, and the attitude of the IMU, on the estimation accuracy of the installation error of the star sensor are also given. Finally, a ten-order rotation scheme has been designed, and all the error parameters can be optimally excited. The simulation and experimental results demonstrate that all the error parameters have better stability and repeatability, when compared with the traditional method, and the initial alignment error of $3.4^{\prime \prime }$ is better than $13.1^{\prime \prime }$ of the traditional method. Moreover, the installation error of the star sensor can be estimated with a high precision only when the angle between the starlight vector and the optical axis of the star sensor exceeds 1°. It is shown that the proposed method can effectively improve the navigation performance of the SINS/CNS integrated navigation system.