Flight taxi-out time prediction method based on queuing theory

Abstract

The invention relates to a flight taxi-out time prediction method based on the queuing theory. The method comprises steps that a static flight taxi-out time prediction model is constructed; actual whole-year historical flight operation data is acquired and pre-processed; historical flights are grouped according to gate position proximity and sliding path consistency principles, the barrier-free taxi-out time of each flight group is calculated; the actual historical flight operation data is utilized, the waiting-for-departure time of a flight is predicted based on the queuing theory; the barrier-free taxi-out time and the waiting-for-departure time are substituted into the static flight taxi-out time prediction model, and model output is the taxi-out time of the flight group. The method is advantaged in that a method integrating theoretical derivation, algorithm implementation, system simulation and case analysis is utilized, the departure flight taxi-out time is predicted by utilizing the static flight taxi-out time prediction model and is predicted based on the queuing theory, and taxi-out time prediction accuracy is improved compared with a traditional prediction method.

Description

technical field [0001] The invention belongs to the technical field of airport operation optimization, and in particular relates to a method for predicting the time of taxiing out of a flight based on queuing theory. Background technique [0002] The taxi-out time refers to the time between the flight taxiing out of the gate and the actual take-off, and is one of the most important indicators reflecting the operational efficiency of the airport. Accurately predicting the possible taxi-out time of the flight before the flight is launched will not only help to improve airport traffic, alleviate taxiing delays, and improve the utilization of taxiways and runways by optimizing the flight launch sequence, but also help to improve flight forecasting. The accuracy of departure time and estimated arrival time is of great significance to airports, airlines and air traffic control. [0003] At present, the research on the prediction of taxi-out time is mainly concentrated in foreign ...

AI Technical Summary

Problems solved by technology

[0003] At present, the research on the prediction of taxi-out time is mainly concentrated in foreign countries, and the representative work is the multiple linear regression model. These studies have analyzed the relevant factors affecting the taxi-out time, and the taxi distance, the number of incoming and outgoing flights and other factors are used as independent variables to construct However, this type of linear regression model based on constant parameters still has considerable limitations in describing the complexity and dynamics of airport operations.

The non-linear method is to use the neural network to predict the flight taxiing time in real time; the flight taxiing process is modeled as a Markov decision process, and a prediction model of the taxiing time based on reinforcement learning is proposed. The average taxi-out time prediction achieved good results, but did not consider the influence of factors such as runway and taxiing distance, resulting in a low prediction accuracy for a single flight; There are two stages of waiting at the end of the runway and queuing at the end of the runway. The first stage uses fixed values, and the second stage uses multiple linear regression to predict. However, for large airports, different airport layouts and taxiing distances will affect the prediction of the first stage. Therefore, setting the barrier-free slide-out time to a fixed value will cause a large error

Images