With the rapid development of the civil aviation transportation industry, the importance of flight safety is becoming increasingly prominent. According to different flight factors and conditions, it is particularly important to model, analyze and evaluate the early warning risk, so as to reduce the probability of flight accidents. This paper builds an early warning model based on aviation safety analysis and flight technology assessment. Firstly, the QAR data was pre-analyzed to determine the variables of data preprocessing, and the variable weight was introduced based on the random forest regression model; second, through the pre-analysis and reference of related literature, combined with the analysis of the pole position change curve were determined to quantify the amount of manipulation of the pilot on the aircraft. Determine the quantification of CAP CLM 1POSN and CAP WHL 1 POSN; next, Based on the frequency number analysis, Over-limit statistics for specific destination airports and over-limit statistics for specific flight stages respectively, Statistics of different overrun characteristics; then, Introduce a classification model based on the gradient lifting tree (GBDT), By repeated reference adjustment, To derive the best parameters of the model, Accurate prediction of the qualification of the landlord control personnel; last, According to the established training model, Overlimit type statistical model, pilot technical evaluation model based on aircraft data, quantitative indicators and QAR data for evaluating relanding based on aircraft data, Establish an automatic early-warning model, Set the threshold value, Determine whether there are any potential safety risks.
Based on aviation safety and flight technology automation early warning research
2023-08-18
2314345 byte
Conference paper
Electronic Resource
English
Civil aviation flight security risk early warning system
European Patent Office | 2020
|