| The development of the civil aviation industry is affected by many factors and has the characteristics of high sensitivity and high risk.In order to effectively avoid risks,it is necessary to take precautions and be keen on the abnormal situation that may occur in the future.Based on this,the early warning methods for the development of civil aviation industry is studied.A civil aviation industry development early warning method based on early warning signal system and Long Short Term Memory(LSTM)is established,which combines the traditional early warning method with the LSTM algorithm in the deep learning method.This method can solve the weaker predictive function of the traditional early warning method and can improve the accuracy of future warning predictions.The research process is divided into five steps: First,select early warning indicators.An early warning indicator system is established through the initial selection and optimization of early warning indicators,which contains 19 three-level indicators;Second,determine the weight of each early warning indicator through the standard deviation correction order relation method;Third,construct an early warning signal system and analyze the alarms of previous years;Fourth,construct the LSTM prediction model and predict the values of the early warning indicators in 2019 and 2020;Fifth,estimate future warnings with the early warning signal system and LSTM predictive models.The research shows that the early-warning method system based on the early-warning signal system and LSTM prediction model can reflect the development status of civil aviation industry better,and the prediction accuracy of early-warning indicators is higher.In 2019 and 2020,the warning situation can be summarized as four points: the growth rate of development scale decreases significantly;The security situation is grim;The problem of insufficient security capability is more prominent;Macroeconomic downward pressure is greater. |
PDF Full Text Request