Adaptive limit checking for spacecraft telemetry data using regression tree learning

This paper proposes an automatic health monitoring method for spacecrafts which adoptively predicts the upper and lower limits of each sensor measurement using a machine learning technique known as regression tree learning. It enhances the widely used limit-checking method so that it automatically and adoptively determines the ranges of numeric variables based on the relationships with relevant symbol variables. We applied the proposed method on the past telemetry data of an artificial satellite and verified its effectiveness.