Mass Property Estimation By EKF and UKF for Spacecraft

The Lunar Gateway is a manned spacecraft being constructed in a near-rectilinear halo orbit around the Moon. Such projects require supply chains in order to sustain human presence and, by extension, a spacecraft which can transport cargo. Unlike common satellites, these spacecrafts have different mass properties for every mission depending on their cargo and can even have different mass properties during the same mission as cargo is loaded and unloaded during docking. Thus, if the mass properties can be estimated on orbit, human resources and fuel can be significantly reduced through the use of effective control.

As a first step toward this goal, this work presents simulation results of a mass property estimation method with the intent to use on orbit. A simulation environment for JAXA’s HTV-X resupply vehicle to the International Space Station was designed. Two methods were tested: an Extended Kalman Filter (EKF)-based method and an Unscented Kalman Filter (UKF)-based method, both of which estimate center of mass and all 6 elements of the inertia tensor simultaneously and in addition to state vectors which consist of attitude and angular velocity.

These two methods were evaluated using simulation data. As a result, the EKF estimated center of mass with less than 1.3 cm accuracy and diagonal elements of inertia matrix with less than 0.8% accuracy. The UKF estimated center of mass within 2.2 cm. accuracy and inertia matrix with 1.2% accuracy. The computational load of the EKF was compared to the UKF and EKF was executed 91% faster. These results demonstrate two robust methods of estimating complete mass properties in the presence of process, measurement, and thruster noise.