Simultaneous Magnetic Dipole and Inertia Estimation for Nano-Satellites Using Dual Filtering

Precise pointing and enhanced attitude stability are increasingly essential for achieving the advanced objectives of nano-satellite missions. For low-Earth orbit (LEO) nano-satellites, the residual magnetic dipole moment (RMM) generally dominates over other environmental disturbance sources, and must be accurately estimated and compensated for precise attitude control. The accuracy of the RMM estimation is highly dependent on the satellite’s inertia tensor, and any deviation in the inertia tensor will compromise the estimation accuracy. This paper presents a dual unscented Kalman filter (DUKF) for RMM estimation coupled with inertia tensor refinement anchored by an inertia prior derived from preflight mass-distribution data. A DUKF uses two separate unscented Kalman filters (UKFs) that are executed alternately within each update cycle, one estimating the RMM and one refining the inertia tensor relative to the prior inertia estimate. As an example, this method is applied to telemetry from the Space Industry Responsive Intelligent Thermal (SpIRIT) 6U CubeSat mission. The resulting analysis demonstrates that the proposed DUKF incorporates inertia uncertainty into RMM estimation and provides a more accurate reconstruction of the observed rotational dynamics compared with a conventional UKF using a fixed precomputed inertia tensor.