Autonomous temperature-based orbit estimation

Abstract Orbit estimation (OE) is a required significant task in almost all space missions. Accordingly, a wide variety of sensors and estimation algorithms have been developed within the last few decades to this aim. However, the current study proposes a novel autonomous OE method that is purely based on temperature data of six orthogonal surfaces of a three-axis stabilized satellite as it orbits around the Earth. While the utility of satellite surface temperature data has been recently investigated for satellite attitude estimation (AE) assuming its navigational information, the present paper is focused on OE via only temperature data that has not been attended to in the related literature. To this end, it is assumed that satellite surfaces are equipped with small plates that are thermally isolated from the internal heat sources so that their temperature changes mainly arise from environmental radiation emanated mainly from the Sun and the Earth. In this sense, a thermal model is developed and demonstrated to show how the satellite surface temperatures and their time rates are the only ingredients needed, as measurement quantities, for the proposed OE method to produce the satellite navigational data in terms of its position and velocity vectors. In addition, the effect of sensor configuration on state observability and estimation accuracy is investigated while the unscented Kalman filter (UKF) is exploited in the estimation process. Performance and viability of the proposed temperature-based OE are verified through Monte Carlo simulations and a comprehensive sensitivity analysis over orbital parameters, satellite initial conditions, sensor accuracy and attitude error.