Onboard Estimation of Mean Orbital Elements Extended to State Jumping Case

Abstract

Mean orbital elements play a significant role in constellation guidance, orbital maintenance, and transfer strategy design. Hence, research on mean orbital elements estimation flourishes with the development of mega-constellation construction, for instance, the Starlink program. This paper investigates the autonomous on-orbit estimation problem of mean orbital elements under the consideration of real-time mean orbital elements acquisition, reliable mean orbital elements prediction, and extreme condition handling of state jumping caused by protective orbit transfer or nonfatal spacecraft collision. A novel mean orbital elements estimator based on the Savitzky–Golay filter (“Smoothing and Differentiation of Data by Simplified Least Squares Procedures,” Analytical Chemistry, Vol. 36, No. 8, 1964, pp. 1627–1639) with adaptive frame size, maneuvering detector, and multimode switching mechanism is proposed. Adaptive frame size according to the orbital period for completing the Savitzky–Golay process guarantees the estimation efficiency and accuracy of mean orbital elements estimation and prediction. Then, the prediction value is employed to generate the mean orbital elements difference compared with the estimation. Finally, the mean orbital elements difference is adopted to govern the satellite state detector for mode switching. Comparison simulations for both with and without state jumping are conducted, and the results demonstrate the superiority of the mean orbital elements estimator over the direct mapping method and the Savitzky–Golay filter in estimation efficiency, accuracy, and range of application.