System Identification with Student’s t-Process Dynamical Model

Abstract

Knowing the exact location of a spacecraft is one of the most important tasks in the operation of a satellite. However, it is difficult to accurately determine the position and velocity of a satellite far from the ground. In order to meet such requirements, various estimation methods have been developed. In this study, we propose a t-process dynamic estimation model, which is based on the t-distribution, and it is suggested that it is more robust than the Gaussian model. In the present study, we have successfully developed a particle filter-based t-process dynamic estimation model.