Inertia tensor estimation of tethered debris through tether tracking

Tethered systems, such as tethered harpoons or nets, are promising methods to capture uncooperative space debris, but suffer from complicated post-capture dynamics and control; control strategies could benefit from knowledge of the mass moments of inertia of the target debris, usually unknown without estimation. In this work, an Unscented Kalman Filter is proposed to estimate the moments of inertia, angular velocity, and attitude of tethered debris in the absence of end effectors and of sensors on the target. This is achieved by leveraging camera measurements of features on the target debris and a tension sensor. Estimation of the properties of (1) a symmetric target with a centered tether attachment point and of (2) a non-symmetric target with a non-centered tether attachment point is performed in simulation. The filter is found to be capable of accurately and precisely estimating both the attitude and angular rates of debris, although sustained loss of a tracked feature heavily impairs these estimates. The principal moments of inertia estimates are unaffected by the loss of a tracked feature and can be made both accurately and precisely when the tether is taut. Asymmetry of the target and of the tether attachment point can have a positive effect on the performance of the filter.