Study On The Model Of Flight Delays Based On Bayesian Network And It's Solutions

Flight delays are the pressing problems of air transportation system in China,causing great economic losses and social discontent. It is urgent but difficult to improve the on time performance of flights, which is still dropping recently. By checking the factors of flight delays thoroughly, and if find the key factors, it is possible to come up with more effective solutions to improve the situation.In this paper, the concept of flight delays is redefined and an analysis on the causes of flight delays is made with reference to the procedures of flight operation.Using the data of Airport X, a model of flight delays based on Bayesian network is built and validated. The model can also be used to forecast flight delays and the causes of them.The structure of Bayesian network and the distribution of posterior probability indicate that, airlines, start flights and aircraft type are the key factors impact on the causes and delays. Some airlines do better than others when their flights encounter unfavorable factors. So airlines should improve their weaknesses and share their superiority to others by marketing. Airlines should also optimize their strings of flights and choose the right type of aircraft to use. Moreover, it is vitally important to guarantee start flights on time. Statistics, incentives and punishment are all helpful to solve the problem of flight delays.