Application of Horizon Sensor Measurements in Kalman Filtering Techniques for On-Board Orbit Navigation During Aerobraking

Abstract

The problem of determining the dynamic state vector, i.e., the position and velocity of a spacecraft as it proceeds along its trajectory, is known as orbit navigation. One approach to this problem is to process a sequence of measurements of the position or velocity through some form of batch or recursive estimation. Numerous approaches to orbit navigation that utilize Kalman filtering of on-board sensors are described in the literature. These sensors include the global positioning system (GPS), magnetic field, star trackers, and horizon sensor measurements. Optical measurements, such as star cameras and horizon sensors, have the advantage of being applied to high-altitude Earth orbits and planetary missions where GPS and magnetic field measurements are not available. In this study, the extended Kalman filtering technique, using horizon sensor measurements, is applied to the problem of real-time on-board navigation during the aerobraking phase of a Mars mission. The technique of aerobraking is used to lower a satellite into its final mission orbit from an initial highly elliptic orbit using successive drag passes at the periapsis to control the removal of orbital energy. The results of Monte Carlo simulations of the inplane orbit dynamics demonstrate the usefulness of the extended Kalman filter approach in the navigation of spacecraft during an aerobraking operation. An improvement in the filter performance is realized when the Kalman filter prototype is augmented with an atmospheric drag force model.