Research on fault diagnosis of high-lift system equipment

The current application level of intelligent technology in aircraft high-lift systems is relatively low, unable to achieve online monitoring and prediction of faults. Particularly, some hidden equipment failures can only be detected when they occur, severely affecting aircraft dispatch rates and flight safety. Focusing on the PDU (Power Drive Unit) motor brake in a specific aircraft high-lift system, this study proposes a method for fault diagnosis that does not require additional sensors. The proposed method is based on forming a database using PDU motor output speed data obtained from test flights and iron bird ATP/QTP tests. By applying the Gaussian Naive Bayesian algorithm for training, a fault diagnosis model is developed and uploaded to the cloud. Utilizing high-bandwidth, low-latency 5G data transmission technology, real-time monitoring data from the aircraft sensors can be transmitted to the cloud-based fault diagnosis model. Cloud computing technology is then used to quickly detect and diagnose the current state of the PDU motor brake and provide early fault warnings. As data accumulates and the confidence level of the algorithm increases, the proposed method can diagnose PDU motor brake faults more effectively and quickly. It offers strong convenience and practicality and can be extended to the fault diagnosis of the entire high-lift system equipment.