LEO Spacecraft Orbit Anomaly Detection Based on Prediction Dispersion.

A detection method based on prediction dispersion was presented to deal with LEO spacecraft orbit anomaly caused by maneuvers, collisions, and atmospheric environment. The semi-major axis and inclination were selected as orbit characteristic parameters. Prediction values of orbit characteristic parameters can be achieved by orbit extrapolation based on SGP4 model secular term. After data smoothing process, prediction dispersion sequence can be obtained by calculating the difference between prediction and catalog values of the characteristic parameters. The orbit anomaly was detected through the prediction dispersion data based on Mahalanobis distance. The detection results of Terra in 2010 is consistent with the maneuver history published by NASA, which indicate that the method presented in this paper can detect the times, epoch and type of orbit anomaly and can be widely applied to the space surveillance and space situational awareness. [ABSTRACT FROM AUTHOR]