Detection and orbit determination of a satellite executing low thrust maneuvers

Concern for the safety of (and risk to) valued space assets has motivated the interest in processes and procedures that enable timely detection of maneuvers for satellites tracked by current and future surveillance systems. The timely detection of maneuvers provides for responsiveness in follow-up tracking, which is crucial for post-maneuver orbit characterization. However, availability and location of surveillance resources may not always allow timely detection and follow-up. Real tracking data for a maneuvering satellite, operated by the Air Force Research Laboratory (AFRL), were used for detection and process performance evaluation. This Low Earth Orbiting (LEO) satellite carried a small thruster that was used for routine orbit maintenance. Range and angles tracking data were obtained over the year 2007. In addition, burn logs for all executed maneuvers were obtained from the satellite operations to support validation of analysis done for this study. Periodic finite burn maneuvers were performed throughout the mission lifetime, sometimes in a sequence separated by hours, providing reference data for use in addressing a variety of maneuver scenarios. The ability to detect and assess maneuvers assuming no a priori maneuver information is examined for single and multiple maneuver scenarios. Batch least-squares (BLSQ) and extended Kalman filter (EKF) orbit determination strategies are applied, analyzed and compared to determine the performance sensitivity to maneuver knowledge. The comparisons examine the reliability in detecting specific maneuvers, the orbit determination performance in processing over the maneuvers, and the subsequent prediction performance resulting from the state estimates. In addition, the techniques presented can also be applied to detecting and supporting resolution of on-board anomalies that might occur on cooperative space assets.