Data mining-based fault detection and prediction methods for in-orbit satellite

Abstract

Fault detection and prediction is one of the key technologies for health monitoring of in-orbit satellites. The fault detection and prediction methods for in-orbit satellites is studied by the data mining technology. The deep mining method for telemetry of the in-orbit satellite is studied. Then, several fault detection and prediction methods are proposed, which have been developed and used in fault detection and prediction of actual in-orbit satellites. The results show that the proposed methods are effective and feasible.