The Research On Autonomous Orbit Determination Of Earth Orbit Satellites Based On UKF

This article is mainly dedicated to the research of the nonlinear filtering in the application to autonomous navigation (auto-Nav) of Earth satellite, and mainly inspects the following aspects: the foundation of satellite dynamics model, the processing of observation data of the sensors and observability analysis of the filtering model, the design of navigation filtering algorithm.Taking the UKF as a basic method, this paper mainly researches the auto-Nav with random observation error only, and quest a high precision technique, suitable for the space environment of auto-Nav. Our research emphasizes on the following four aspects:1. The research on the development of auto-Nav and nonlinear filtering By the research of auto-Nav method and technique, we summarize the one that fits for the reality of our country from the tendency of auto-Nav. By measuring various filtering, we conclude the one that can fit the complex environment very well.2. The research on the influence of the perturbing forces on the orbit From the satellite orbit dynamics, we do some research on the main perturbing forces, and get to the conclusion of influence of perturbing forces on orbits.3. The research on observability analysis of nonlinear systemsTo the nonlinear system made up of the state equation and the observation equation, we apply the principal of observability of linear systems to this one, and get the observability judge matrix of nonlinear systems.4. The research on auto-Nav with UKFFor two observation equations with random error only and constant observation error, we apply UKF to the auto-Nav, and conclude that it's feasible for the method. Through deep research on the model with random error only, we know the main factors that lay influence on the auto-Nav precision. And design the navigation filtering algorithm from the above analysis and simulation.The main problems in auto-Nav are observability and the precision, i.e., feasibility and reliability. The observability analysis results indicate that the method is feasible. The simulation results indicate that the precision of auto-Nav is very good, and robust to the space environment.