Fault diagnosis of attitude control sensors for drag-free satellite with single test mass

A fault diagnostic method based on a set of extended Kalman filters (EKFs) is proposed for sensor faults in an attitude control system of drag-free satellite with single test mass. Firstly, the system model of a drag-free satellite with a single test mass is established, and the gravitational reference sensor modelunique to the drag-free satellites is further proposed. A set of EKFs is designed based on the model of a dragfree system with a single test mass to estimate the system state, including a total filter and four parallel subfilters that receive different types of sensor data. Subsequently, a multi-sample residual chi-square test method is employed to analyze the estimation residuals generated by the filters, enabling sensor fault detection and preliminary fault isolation. Finally, the specific fault locations of fault sensors can be further isolated by calculating and comparing the values of the generalized likelihood function based on the estimation residuals. Simulation results confirm the effectiveness of the proposed method.