Research On The Impact Of Severe Weather On Flight Delay

Delays cause transportation services waste of airlines and the society,and even harm the social public order.Bad weather is one of the main reasons for delays,which may cause massive delay even flights cancellation.With sophisticated ATC technology and improvement of the performance and maintenance ability of modern aircraft,meteorological conditions have become the main causes of the punctuality and safety of civil aviation.By lowering the available resources of airports and air routes,adverse weather causes flight delays and even endangers the safety of civil aviation.Bad weather occurs at random,therefore making it more difficult to research its uncertain influence.This paper explores how to integrate and effectively use the existing plenty of historical data and explore its relationship with delay.This project has important practical significance for improving decision-making level of each civil aviation operation departments,but also has important academic value for theoretical research on flight delay.This paper firstly expounds weather effect on flight delay in detail,and the definition of bad weather.Secondly,we classified the propagated delay caused by former fight segment and independent departure delay.Meteorological data analyzed and extracted from METAR,is combined with corresponding traffic flow data.And on this basis this paper build different models.The average delay time of single OD pair is build with C5.0 Decision Tree model.This paper also applies boosting technology,pruning technology and cost matrix to further optimize the model.And a case study of ZBAA-ZSPD delay time prediction is given.The precision of the model reaches 72.8%.The average delay time of main OD pairs are build with BP neural network model,which contains 6500 input nodes,380 output nodes.The simulation results show that the model is in consistent with requirements.