Departure capacity prediction for hub airport in thunderstorm based on data mining method

Abstract

Capacity prediction is significant for flight plan in airport, which can contribute to the reduction of flight delays and passenger backlogs, thus ensuring the orderly operation of hub airport in thunderstorm. The factors such as weather and traffic control often lead to a large-scale flight delays. Thunderstorms is the major factor that results in the decline of airport operation capacity, which has the characteristics such as difficult to predict, dynamic evolution and large sphere of influence. This paper analyzed the factors which affect the departure capacity according to the operation data of the past five years in Beijing Capital International Airport, then established a prediction model for the departure capacity based on Support Vector Machine (SVM) and developed software installed in the operation control department. The results indicated that the prediction model has practical value with relatively high accuracy. And it can be concluded that this capacity prediction model can reduce the flight delays and ensure that the number of flights matches the support capability of the operation control of Beijing Capital International Airport.