Kernel-based ensemble gaussian mixture filtering for orbit determination with sparse data

In this paper, a modified kernel-based ensemble Gaussian mixture filtering (EnGMF) is introduced to produce fast and consistent orbit determination capabilities in sparse measurement environment. The EnGMF based on kernel density estimation (KDE) combine particle filters sum filters. This work proposes using Silverman’s rule of thumb reduce the computational burden KDE. Equinoctial orbital elements are used improve accuracy KDE bandwidth parameter EnGMF. A bi-fidelity approach propagation an adaptation algorithm for selecting appropriate number particles also applied with acceptable loss long time propagation. Through numerical simulation, proposed implementation compared state-of-the-art approaches terms accuracy, consistency, speed.