Skip to Main Content
Principal component analysis (PCA) combining with multivariate statistical knowledge is used for the sensor fault detection and diagnosis according to the characteristics of the satellite attitude control system. In this paper, the principle of PCA to detect faults is presented, and the conventional PCA fault isolation approach is improved. The example of using PCA to fault detection and diagnosis of the typical fault of the infrared earth sensor is given, which is based on faults simulation. The result shows that it is feasible for the fault diagnosis of sensors in the satellite attitude control system and the PCA approach has good performances in fault detection and diagnosis.