Reliable Kalman Filtering for Satellite Attitude Estimation Under Gyroscope Partial Failure

摘要

A147 estimation is one of the important aspects in Attitude Determination and Control System (ADCS) of a satellite. Under normal working conditions, it is possible to achieve sufficiently a good attitude estimation using regular Kalman filtering. However, in the presence of sensor faults, the filter fails to provide the desired accuracy and may deteriorate the attitude estimation. To overcome this imperfection, a Fault Tolerant Extended Kalman Filter (FTEKF) is presented in this paper to improve the state estimation reliability even in the sensor fault situation. In accordance, a model of sensitivity factor is applied in the fault detection design of the filter to detect gyroscope sensor faults. Numerical comparison with Robust Extended Kalman Filter (REKF) is carried out for performance assessment. The proposed filter is approved to be robust against gyroscope faults with better filtering performance.