Heavy Landing Detection and Prediction Based on QAR Data

The approach and landing phase is the high accident phase throughout the flight, and the most typical accident is heavy landing. Heavy landing will not cause too many casualties, but it will cause damage to the aircraft structure and increase flight safety risks. Therefore, this paper uses a clustering algorithm to filter the key parameters in the landing phase, and then combines the Autoregressive Integrated Moving Average Model (ARIMA) with the Support Vector Machine Model (SVM) in series. The former carries out parameter prediction, the latter carries out heavy landing detection to jointly realize the real-time warning of heavy landing. It is of great practical significance to avoid heavy landing events by operation in time in the actual flight process.