Cubesats provide an inexpensive platform on which to place small payloads in space. However, the small volume of cubesats limits the size of each subsystem. Ideally, a payload requiring fine pointing accuracy would utilize very accurate attitude determination sensors, such as star trackers, but such sensors generally have large footprints. Therefore, it is desirable to develop an attitude estimator using a set of low-profile sensors. Three Extended Kalman Filters are developed for a cubesat with a limited sensor suite consisting of a MEMS magnetometer, a MEMS gyroscope, and an array of single-axis sun sensors. The first filter estimates attitude and attitude rate using the full sensor suite. The second filter estimates attitude and attitude rate without gyroscope information. The third filter estimates attitude, attitude rate, and gyroscope bias using the full sensor suite. All three Extended Kalman Filters use the backward-difference time derivative of the measured magnetic field as an additional measurement. An analytical time derivative of the measured magnetic field is developed and used in the nonlinear measurement model.
展开▼
