Satellite online fault diagnosis method based on data mining

A method for online fault diagnosis of satellites based on data mining is demonstrated. The present study is divided into two parts: online detection and automatic fault diagnosis. First, preprocessing of telemetry data is performed. Then, abnormal data detection methods are abstracted to form a set of monitoring methods with standardized representation forms, achieving online anomaly detection for satellite status. Once an anomaly is detected, modeling and formalized representation of fault diagnosis logic application models are performed, followed by fault localization diagnosis, thereby enabling rapid diagnosis and analysis positioning of faults in satellites and other large-scale complex equipment.