Application of Constrained Filtering to LEO Satellite Attitude Determination

• Main Authors: Yi Chen, 陳逸
• Other Authors: Doctor Fan-Ren Chang
• Format: Others
• Language: zh-TW
• Published: 2002
• Online Access: http://ndltd.ncl.edu.tw/handle/83651784960899214701

碩士 === 國立臺灣大學 === 電機工程學研究所 === 90 === The main topic of this study is about the attitude determination systems of satellites. First, the concepts of coordinate systems and coordinate transformations are introduced, as well as the kinematics and dynamics of satellite attitude motion. We then discuss two types of attitude sensors, the magnetometer and the earth sensor, which are widely used on LEO（low earth orbit）micro-satellites to measure the satellite attitudes. In addition, we describe the operating principles and usages of these two sensors, the mathematical models of them, and the applications of the sensors to attitude determination. Two filtering methods are proposed to fulfill the task of attitude determination. On the one hand, we describe the concepts of maximum likelihood estimation and constrained filtering method, and explain how to apply them to the problem of satellite attitude determination. On the other hand, when researchers are going to deal with the issues of orbit position or attitude of a satellite, conventionally, an effective solution is to use the extended Kalman filter to track the system state variables. For comparison, we also design an extended Kalman filter to deal with the same attitude determination problem. To integrate all the ideas mentioned in the study, we also scheme out some numerical simulations to test the proposed sensor models and the filtering algorithms. The simulation results demonstrate the reliability and effectiveness of the attitude determination system in our study.