TRIAD-Aided Multiplicative EKF for Small Satellite Attitude Estimation and Magnetometer Calibration

This study presents a nontraditional estimation system for attitude estimation and complete magnetometer calibration of a small satellite. The proposed system combines the TRIAD algorithm, a static attitude determination method, with a multiplicative extended Kalman filter (MEKF) to improve accuracy, and enhance system flexibility. The TRIAD algorithm serves as the initial step, providing an initial coarse quaternion set estimation using measurements from a three-axis magnetometer and a sun sensor. Subsequently, the coarse estimation is filtered through the MEKF to obtain the final estimation. In addition to attitude estimation, the MEKF performs magnetometer calibration by estimating biases, scaling factors, and nonorthogonality corrections. To evaluate the performance of the proposed system, multiple numerical simulations are conducted using a hypothetical nanosatellite. The simulation results include estimations of attitude, gyro bias, and magnetometer calibration parameters, as well as an assessment of the system's performance under various time-varying magnetometer biases.