Exhaust Gas Temperature Sensing Data Anomaly Detection for Aircraft Auxiliary Power Unit Condition Monitoring

Auxiliary Power Unit (APU) is one of the most subsystems in the civil and military aircraft. Its reliability is directly relative to the aircraft availability. Condition monitoring can be utilized as one effective method to enhance APU reliability. Among the practicable techniques, data-driven condition monitoring can be achieved by using the on-wing sensing data of APU. Among the on-wing sensing data, Exhaust Gas Temperature (EGT) is one of the most important parameters. In this study, EGT is detected to ensure whether it is anomalous. The anomaly detection approach is based on Relevance Vector Machine (RVM) which has the sparse feature. The historical EGT sensing data are adopted to train RVM model. Then, the real-time EGT sensing data are detected. For experiment evaluation, the on-wing EGT sensing data of APU from China Southern Airlines are utilized. The effectiveness of the proposed approach is proved.