Roll estimation for smart munitions using a 3D magnetometer only

Accurate roll angle and roll angular rate are important to improve the target detection accuracy for the smart munitions. Because of the limited space in the fuze, only a three-axis magnetometer is used to estimate the roll attitude on the whole trajectory. The two methods of roll estimation are proposed and compared. One is an iterative method based on the attitude transformation matrix, in which the pitch angle and yaw angle at the previous moment are utilized to estimate the roll angle at the moment and then the three angles are kept being corrected. And the other is an extended Kalman filter (EKF) method. The state space model is derived by the dynamic characteristic of the smart munitions, and the measurement model is derived by the relationship between geomagnetic field and the output of a three-axis magnetometer. The simulation results show that the EKF method is able to accurately estimate the roll attitude and is more practical than the iterative method.