In-flight alignment of inertial navigation system by celestial observation technique

This paper presents an in-flight alignment technique for a strapdown inertial navigation system (SINS) and employs a star pattern recognition procedure for identifying stars sensed by a CCD electrooptical star sensor. Collinearity equations are used to estimate sensor frame star coordinates and the conventional least square differential correction method is used to estimate the unknown orientation angles. A comparison of this attitude with the attitude estimated by the SINS provides axis misalignment angles. Simulations using a Kalman filter are carried out for an SINS and the system employs a local level navigation frame. The space stabilized SINS is discussed in conjunction with the celestial aiding. Based on the observation of the Kalman filter, the estimating and compensating gyro errors, as well as the position and velocity errors caused by the SINS misalignments are calibrated by celestial attitute information.