Initial Alignment for a Doppler Velocity Log-Aided Strapdown Inertial Navigation System With Limited Information

This paper proposes an ingenious alignment scheme for a Doppler velocity log-aided strapdown inertial navigation system with only the initial velocity and position information. The proposed scheme decomposes the task into two tightly coupled subproblems. First, an optimization-based coarse alignment (OBCA) method is developed to derive the constant attitude at the very start of the alignment. During the OBCA process, the required velocity and position are all approximated using their initial values. Second, a nonlinear-filtering-based fine alignment (NFFA) is developed to refine the attitude and calculate the real-time velocity and position. In an NFFA, the standard inertial navigation equations are used directly as the process model, which, therefore, can track the vehicle maneuverability during the alignment period. The NFFA is carried out again from the very start, initialized with the roughly known attitude by OBCA and the known initial velocity and position information. This scheme necessitates the recording of the sensor data during the OBCA. The experimental results show that the proposed method can not only align the attitude to high precision within very short time, but also determine the velocity and position with satisfied accuracy.