Qauternion-based attitude filtering and on-board sensor calibration for small satellies

Abstract

This thesis presents a comprehensive analysis and novel methodologies for quaternion-based attitude filtering and on-board sensor calibration, specifically tailored to meet the needs of small satellites. It focuses on developing various attitude estimation strategies to reduce computational costs and enhance fault tolerance. Additionally, it encompasses comparative analyses between various algorithms and tackles the challenges associated with sensor calibration. The research, primarily based on simulations using MATLAB and confined to the independent attitude determination of a single small satellite in Low Earth Orbits (LEO), excludes actual sensor data and flight formation scenarios. Emphasizing quaternion methodologies over others, like Euler angles, it doesn't delve deeply into related areas such as attitude control or orbital mechanics. The thesis proposes innovative attitude estimation systems integrating single-frame methods with Kalman filtering for improved convergence and adaptability, and includes gyroscope and magnetometer calibration within the state vector for enhanced accuracy. Targeted at practical implementation in satellite systems, especially where hardware redundancy is limited, the thesis also aims to serve as a comprehensive academic resource for future research and development in small satellite attitude estimation.