Spacecraft angular velocity estimation using sequential observations of a single vector

A new approach is presented for estimating the angular rate of a tumbling, momentum wheel-equipped spacecraft from sequential measurements of a single directional vector. The method consists of a two-stage algorithm. The first stage is a deterministic algorithm that provides a coarse estimation of the satellite angular velocity. Combining the kinematics equations, corresponding to the apparent motion of the vector measurements in the satellite body frame, with the satellite dynamic equations of motion, enables the development of an almost one-to-one relation between the vector measurements and the satellite angular rate. This relation is used to reconstruct the satellite angular velocity. The second stage is an extended Kalman filter which is initialized using the first stage crude estimates. This combination renders the filter accurate and robust. A simulation study, in which angular rates of up to 0.5 rad/s were implemented, is used to demonstrate the performance of the method.