摘要

In this work, the main objective is to study a real-time H-infinity cubature Kalman filter (CKF). Robust CKF combines robust principle and CKF, many advantages in state estimation for a small unmanned helicopter are presented, employing singular value decomposing. Meanwhile, a real-time principle is introduced to overcome the problem of costly calculating time. Unlike many other novel CKF is verified in simulation, the proposed CKF is implemented on the small scale unmanned helicopter's inertial navigation system. A serial of experimental results on a small-scale helicopter confirm the effectiveness scheme and the accuracy state estimation of the proposed filter in high-maneuverability flight.