Prediction model of GEO spacecraft position after maneuver based on causal Bayesian network

Predicting the position of Geosynchronous (GEO) spacecraft under maneuver is a crucial work for space domain awareness (SDA) as it can help to improve the flexibility and operational efficiency of space surveillance network (SSN). The longitude is a unique freely assigned parameter for GEO spacecraft, in this paper, a predictive model of GEO spacecraft longitude based on causal Bayesian network is proposed. Firstly, the causal parameters of longitude is found by Gaussian perturbation equation. Secondly, the Markov order of the causal parameters is obtained by transfer entropy. Finally, the linear expressions for Gaussian functions is proved and a causal Bayesian network (CBN) prediction model for longitude is constructed. After experimental analysis, the mean absolute error (MAE) and mean square error (MSE) of the proposed method are decreased by 71.88% and 72.18%, respectively compared with the traditional Long short-term memory (LSTM) method.