研究目的
Investigating the performance of a quaternion-based attitude unscented Kalman filter in a magnetometer-only spacecraft scenario.
研究成果
The quaternion-based attitude UKF using small angle approximations is robust, accurate, and rapidly convergent for small true gyro biases and small initial uncertainties. However, its performance is highly dependent on the initial gyro bias covariance and true gyro biases, making it less robust than an EKF in some scenarios. The study suggests that the EKF's superior robustness and the UKF's faster convergence under optimal conditions highlight the need for careful selection of filtering approaches based on specific mission requirements.
研究不足
The UKF's performance depends significantly on the initial gyro bias covariance and true gyro biases. It is less robust than an EKF with the same attitude approach under certain conditions. The study is limited to a magnetometer-only scenario, which may not fully represent all operational conditions.
1：Experimental Design and Method Selection:
The study uses a quaternion-based attitude unscented Kalman filter with small angle approximations for quaternion errors. The state vector includes the attitude quaternion and gyro bias vector. The filter is evaluated in a simulated lost-in-space scenario of a low-Earth orbiting spacecraft using only three-axis magnetometer and gyro measurements.
2：Sample Selection and Data Sources:
Simulated data of NASA’s Tropical Rainfall Measuring Mission (TRMM) spacecraft is used, with extensive Monte Carlo runs to evaluate performance metrics like robustness, accuracy, and speed of convergence.
3：List of Experimental Equipment and Materials:
The study involves a simulated three-axis magnetometer and gyro measurements, with noise characteristics specified for both.
4：Experimental Procedures and Operational Workflow:
The filter's performance is evaluated under various initial conditions, including different true gyro biases and initial gyro bias covariances, to assess convergence and accuracy.
5：Data Analysis Methods:
Performance is analyzed based on convergence times, attitude accuracy, and comparison with other Kalman filter approaches (EKF and another UKF using generalized Rodrigues parameters).
独家科研数据包，助您复现前沿成果，加速创新突破
获取完整内容
