Modeling and semi-physical simulation of in-flight alignment for MIMU under large heading error

The micro inertial measurement unit (MIMU) in micro unpiloted aerial vehicles initial alignment time is short, so in-flight alignment for MIMU is worked under large heading error with the global positioning system (GPS) assistant. Aimed at the requirement of high estimation precision and real time, a linear model with large heading error is introduced by regarding nonlinear vector as state vector, and adaptive Kalman filter (AKF) base on this linear model is used for MIMU in-flight alignment. The semi-physical simulation results show that this method has better precision than extended Kalman filter (EKF) with nonlinear model, and the heading error descends very quickly from 60° to around 2°. Furthermore, AKF need less computation time than EKF, only 67% of EKF.