Enhancing Simplified General Perturbations-4 Model for Orbit Propagation Using Deep Learning: A Review

This paper studies the method used in orbit propagation in order to enhance the Simplified General Perturbations-4 (SGP4) model which is the common orbit propagation model used by the satellite operator. The orbit propagation is used to determine and predict the position and velocity of a satellite. The capability of making an accurate orbital prediction is important to ensure satellite operation planning will not be disrupted and prevent any disrupted collisions or disasters. However, the accuracy of the SGP4 model is decreased once the propagation horizon increased. Therefore, a study is done to identify a technique that can be applied to enhance the SGP4 model. The model needs to be improved in term of minimizing the error and increase the accuracy even though the propagation span is increased. The method used in this study is by comparing the techniques that have been used by other researchers for the orbit propagation model. From the review that has been done, a deep learning technique is found to be a suitable technique. It also produces an accurate model for time series data. The new framework of the SGP4 model is expected and able to become a reliable predictor model. In the future, the study will further analyze by using simulation tools and real-time data. The accuracy and effectiveness of the improved model will be evaluated and the results will be compared with actual observations.