Application of Adaptive Kalman Filter Model Error System in Initial Alignment of Strapdown INS

For the inertial navigation system with model errors,there exists a large error in estimating a state using conventional Kalman filter,and it can even make the filter diverging.This paper presents a new method,which employs adaptive Kalman filter algorithm.It improves the design of Kalman filter online by introducing the dummy noise and using the observation information,and apply it to the initial alignment of strapdown inertial navigation system.The estimation accuracy of the filter using this method is higher than that of conventional Kalman filter.The examination and the simulation show the validity of this method.