Research on Unstable Approach Detection of Civil Aviation Aircraft

Abstract In order to avoid serious civil aviation accidents caused by low altitude and unstable approach, the empirical model of unstable approach below 1000 feet is established by means of pilot survey. By using AGS decoding platform, unstable events are triggered. With Python, the real values of trigger time are extracted and stored and then compare and analyze the data. Experiments are conducted on 30,000 actual QAR data of B-737 aircraft. The results showed that 3100 unstable approach segments were triggered, and the ratio was about 10.3%. Two of the main factors causing unstable approach are the overrun of air speed and the deviation of localizer. Through the study and analysis of various factors of model, preventive measures can be taken in order to reduce the incidence of unstable approach.