Machine learning for atmospheric drag prediction of LEO satellites

In this study, a new atmospheric drag prediction method for spacecraft using machine learning is proposed. In the proposed method, a machine learning model is constructed with the orbital decay rate of each spacecraft and the space environment factors having a strong influence on the upper atmosphere. To demonstrate the effectiveness of the proposed method, the analysis of two satellites, ALOS-2 and GCOM-W1, is conducted. The results show that the proposed method based on machine learning can predict the atmospheric drag with a relatively good accuracy.