Abstract:

A passiveBD/serial inertial navigation system (SINS) integrated navigation system using traditional indirect Kalman filtering method can not achieve the high positioning accuracy and costs too much calculating time. An improved filtering algorithm based on unscented Kalman filter (UKF) is proposed after comparing the advantages and disadvantages of the traditional direct filtering method, the extended Kalman filtering (EKF) method and UKF. Navigation parameters from SINS and platform error angles are chosen as the state of the system, and navigation parameters from passiveBD are chosen as measurements. The final optimal estimations of navigation parameters including position and speed can be calculated directly. The experimental result indicates that the new algorithm requires no accurate error models and can effectively solve the nonlinearity problem of state equations. Besides, it can also simplify the adjustment process of filtering parameters to reduce the total calculating time.  Comparing to prevalent filtering algorithms, this new one can obtain more accurate positioning result and is significantly important to the further study of the filtering algorithm for passiveBD/SINS integrated navigation system.