Hybrid data-driven anomaly detection method to improve UAV operating reliability

The paper presents a hybrid data-driven approach of anomaly detection for UAV (Unmanned Aerial Vehicle) system. Specifically, it is focused on implementing on-line abnormal discovery to improve the operating reliability of UAV. The anomaly detection framework is based on time series segmentation, associated rules mining and associated anomaly detection. Experimental results through simulation and actual flight data, demonstrate that anomaly detection and identification with the monitoring sensor data can be conducted for UAV system.