Spacecraft inertial parameters estimation using time series clustering and reinforcement learning

TL;DR

This paper introduces a machine learning method combining time series clustering and reinforcement learning to estimate and adapt spacecraft inertial parameters during dynamic operations, improving accuracy and resilience.

Contribution

It presents a novel approach integrating time series clustering with reinforcement learning for real-time inertial parameter estimation in changing spacecraft conditions.

Findings

Method effectively distinguishes different inertial parameters.

Algorithm shows resilience to operational disturbances.

Performance validated on multi-satellite deployment scenario.

Abstract

This paper presents a machine learning approach to estimate the inertial parameters of a spacecraft in cases when those change during operations, e.g. multiple deployments of payloads, unfolding of appendages and booms, propellant consumption as well as during in-orbit servicing and active debris removal operations. The machine learning approach uses time series clustering together with an optimised actuation sequence generated by reinforcement learning to facilitate distinguishing among different inertial parameter sets. The performance of the proposed strategy is assessed against the case of a multi-satellite deployment system showing that the algorithm is resilient towards common disturbances in such kinds of operations.