A Review of Flight Delay Prediction Methods

With the rapid increase of civil aviation transport volume, the problem of flight delays is still extremely serious. Effective prediction of flight delays will not only reduce the economic losses caused by airline delays, but also reduce the stay time and improve the traveling satisfaction of passengers. First of all, this paper describes the various time nodes of the complete flight operation process in detail. Next, it systematically expounds the reasons for the generation and propagation of flight delays, and puts forward two solutions, i.e. flight cancellation and aircraft swapping. Then, the main characteristics of flight delay prediction methods are summarized into three categories: traditional statistical analysis, modeling simulation and queuing theory, and machine learning methods. In addition, we analyze the principles and implementation processes of the representative method in detail, and compare their advantages and disadvantages. Finally, this paper proposes the future research direction of flight delay prediction methods.