In view of the low precision of microelectromechanical systems(MEMS) gyroscope and the inability to complete the azimuth alignment of the inertial navigation system, a method of using the accelerometer and magnetometer output to complete the coarse alignment of the MEMS inertial navigation system is proposed in this paper. At the same time, aiming at the shortcomings of the forgetting Kalman filtering method, an alignment scheme of introducing the multi forgetting factors adaptive Kalman filter (MFFAKF) method into the system is proposed. The method uses the index fading memory means to calculate the variance matrix of measurement noise, and the multi forgetting factors is used to adjust the different filtering channels. The experimental results show that the coarse alignment heading accuracy obtained by MFFAKF satisfies the coarse alignment requirements of the MEMS inertial navigation system.