Neural Network Based Orbit Propagation for Small Satellite Missions

The Global Positioning System (GPS) receivers are being widely utilized for low earth navigation in space. These receivers provide the orbital information in real time while the attitude sensors provide real time attitude data. For full autonomy of a satellite, any failures which may occur due to different sources, such as GPS receiver, attitude sensors are to be detected autonomously and suitable remedial actions need to be taken to continue the satellite operations. Towards achieving autonomy, GPS receivers are widely used onboard the satellite for orbit information. In case of a failure or outages of the GPS receiver data, an Artificial Neural Network (ANN) based back-up scheme is proposed in this research to handle these failures in orbit determination. ANN has the advantage of no requirement of system modeling. They are employed in this research to detect attitude and orbit sensor failures in view of their good performance and flexibility in implementing onboard a LEO small satellite