Research On Aircraft Departure Dynamic Pushback Control Method Based On Predicted Taxi-out Time

Air transportation industry is an significant part of comprehensive transportation system.The ability of air transport develops in high-speed with the increase of aviation business scale in recent years.Consequently,the bottle of airport capacity is emerged and the experiential management mode has not satisfied the departure traffic operation demand in current infrastructure.Along with a lot airport surface congestion and flight delay,which not only can deduce the airport operation efficiency,but also make airlines and passengers suffer enormous economic loss,as well as the gas emission.In order to solve these problems,scholars make a lot of studies mainly focusing on taxiing time estimation,taxiing routes optimation,pushback sequence optimazion.However,study on aircraft departure pushback control method is very few,as well as pushback control dynamically.This thesis using departure process as research carrier,pushback process as object,build the aircraft dynamic pushback control model based on taxi-out time prediction.The objective of this research is curtailing the departure operation cost and surface congestion.The departure surface operation process and the influencing factors of the departure taxi-out process are described.The departure departure data in PEK airport are screened and statistically analyzed.Based on this work,the well-known N-control pushback control method is described and simulated.Based on this,a penalty-based dynamic pushback control method(DPC method)is proposed in which the pushback rate changes with the current queuing length of the taxiway in real time.The mathematical model of the method is established,and the Markov process of the model is analyzed.An iterative continuous timeMarkov chain-based optimization algorithm is designed to solve the model.In order to verify the mathematical model of the method,a DPC method simulation process of the departure process is designed.On the basis of the constant taxiway queue length threshold,a time epoch based DPC method is proposed,which makes the taxiway queueing threshold change according to the traffic volume in different time periods.Four scenarios of airport closure are designed to analyze the applicability of pushback methods.The main influencing factors are analyzed and the taxi-out time prediction evaluation indicators are proposed.The impeded taxi-out time is defined,predictive model of which is established.According to actual departure taxi-out time,four predictive models based on regression technic,which are Linear regression prediction model,Generalized linear regression prediction model,Softmax regression prediction model and Support Vector Machine regression prediction model.Moreover,two regressive predictive model based on intelligent algorithm optimization,which are Support Vector Regression prediction model based on Particle Swarm Optimization algorithm and Support Vector Regression prediction model based on improved firefly algorithm,are build.The predicted results of each predictive model are compared and analyzed.The results show that IFA-SVR model can achieve 79.39% and 95.52%,respectively,on predictive accuracy within 2min and 5min.Through screening the taxi-out time prediction approach and dynamic pushback control method,a predictive taxi-out time based dynamic pushback control method(PDPC)is proposed.The control model is built and the solving alogrithm of which is designed.The PDPC method is used to simulate the departure process and the operational benefits of the airport are analyzed in terms of the number of pushback aircraft,the number of taxiway queuing length,and the operating cost per unit time.Ther results show the PDPC method can achieve a 36.25% reduction on total operational cost than no pushback method situation.Four airport closed scenariosunder extreme conditions,which are “off-peak periods close,off-peak perids open”,“off-peak periods close,peak perids open”,“off-peak periods close,peak perids open” and “peak periods close,peak perids open” are designed to analyze the practicality of the PDPC method.Finally,with the objectives of simple operation mode and optimal cost,three kinds of departure hybrid dynamic pushback control policies,which are N-control-PDPC control policy,Step function control policy and Linear-nonlinear control policy,are proposed,modeled and simulated.The simulation results of each policy are compared with their respective application characteristics.This thesis studies the predictive models and algorithms of departure taxi-out time and dynamic pushback control policies to enhence the accuracy of airport operation control and improve the experiential management mode.At the same time,it can alleviate airport scene traffic congestion,reduce taxi fuel consumption and waste emissions,as well as bring economic benefits to airports,airlines and passengers.The research results in this thesis provide new ideas for aircraft control scheduling,decision-making and optimization,and have certain reference value for increasing the relative capacity of airports and establishing an intensive airport dispatch management model.