Congestion Identification And Prediction Of Air Traffic Network Flow System Under Large-scale Flight Delays

Large-scale flight delay problem has always been a big problem,the civil aviation transportation industry with the rapid development of industry,the problem of large-area flight delays,bad weather,airspace is busy,poor environmental conditions,such as airport facilities failure can cause large-area flight delays,resulting in air traffic congestion in network flow system.The identification and prediction of congestion is the basis of traffic management.This paper mainly studies from the following aspects: congestion identification,congestion degree assessment,congestion prediction and congestion degree prediction:Firstly,based on the concept of air traffic network flow system,the definition of large area flight delay and congestion is given.The existing quantitative indicators and classification are given,and the characteristic attributes of congestion,such as time,place and cause,are analyzed,which provides a theoretical basis for the following research.Then,the status of air traffic network flow system is divided into three types:normal,congestion and congestion,and the degree of congestion is subdivided into slight congestion,congestion and severe congestion.A congestion identification model is established based on the congestion index to judge whether the system is congested.The congestion degree evaluation index system is established based on the airport and airline,the entropy method is used to calculate the index weight,the congestion degree evaluation model based on the fuzzy comprehensive evaluation method is established,and the calculation example is analyzed respectively.It provides scientific basis for quantitative description of congestion.Finally,the congestion prediction model of BP neural network optimized by genetic algorithm is established,and the congestion identification algorithm is combined to predict whether the system and the airports and airlines within the system will be congested in a certain period of time in the future.A prediction model of congestion assessment parameters optimized by BP neural network based on genetic algorithm was established.Combined with the congestionassessment model,the congestion degree of the system and all airports and airlines in the system was predicted.The numerical examples are analyzed respectively.