Multiplicative Quaternion Extended Kalman Filtering for Nonspinning Guided Projectiles.

摘要

The problem of estimating the attitude, velocity, and position states of a slow to nonspinning, gun-launched projectile is addressed through the use of an extended Kalman filter (EKF). The EKF is constructed using an error-state mechanization similar to those used in traditional inertial navigation applications, although the kinematics are simplified by assuming the Earth-fixed reference frames used for tactical applications are inertial. The advantages of using quaternions rather than Euler angles to represent projectile attitude are discussed, and the use of multiplicative quaternion error states is given detailed attention as this leads to fundamental differences from most other extended Kalman filter implementations that tend to assume additive error states. The measurements and heuristic information available for most projectile applications are incorporated into the EKF. Two sets of simulation results involving a direct-fire system and an indirect-fire system are presented.