Research On Traffic Operation Status Identification And Prediction Technology Of Air Route Network Based On Flight Trajectory Data

As an essential component of the air traffic management system,the free flow of the air route network,or the lack of it,hugely impacts the safety and efficiency of the air traffic management system.The construction of the intelligent ability to percept the operating status of air traffic circulation plays a fundamental and essential role in comprehensively developing a smart air traffic management system.It enables the air traffic management departments to fully control the overall operating status of the air route network,analyze the bottleneck and critial node of traffic circulation,establish reasonable and optimized management measures,and realize the sophisicated management of air traffic.It helps airlines to optimize their flights,minimize flight delays,and reduce flight costs.The current air route network traffic circulation is under an extensive managament based on tradition and experience,and without clearly-targeted and sophisicated management and decision-making,the operating efficiency of the air traffic system is largely reduced,and thus poses a negative influence on the flight on-time performance level to some extent.Currently,research on the identification and prediction of air route network traffic status is limited,and a scientific theory is not formed yet.Based on the flight trajectory data of aircrafts,this thesis seeks to lay a theoretical and technical basis for the construction of a smart perception system of air route network traffic status through profound and systematic studies on the analysis of air traffic network circulation features and the evaluation,identification and prediction of the traffic status of the air traffic network and its segments.(1)An analysis on the space-time characteristics and evolution law of the traffic flow circulation in the air traffic network.Based on the flight trajectory data of the air traffic network,the thesis analyzes the temporal and spactial properties of the traffic flow circulation in the air traffic network.In terms of temporal property,the change of traffic flow is found to demonstrate a chaotic charactieristic,a power law relation between the mean value and the variance of the air traffic flow volume is verified,and the fluctuation of the volume is influenced by both internal and external factors.In terms of spatial property,a spatial relativity is detected in the change of traffic flow,and its degree of intensity is later used in the group division of traffic segments with multidemensional scaling method,which supports the prediction of multi-segment traffic flow parameters.(2)Evaluation and division methods of the circulation status of the air traffic network and its segments and their correponding models.Based on the flight trajectory data of the air traffic network,an evaluation model on the overall traffic circulation status is established using fuzzy comprehensive evaluation.A division model on the traffic cieculation status of air traffic segments is established based on SAGA-FCM to realize the fuzzy division of the overall network and its segments and set the method and standard of classifying the traffic circulation status.(3)An identification method of the circulation status of the air traffic network and its segments and its correponding model.Based on the mothod of machine learning and pattern recognition,the identification model of the overall traffic circulation status is established using SVM,and the identification model of the traffic segments circulation status is established using ensemble learning.Based on the flight trajectory data of the air traffic network,an identification method of the traffic circulation status is presented to support the prediction on the circulation status of the air traffic network and its segments.(4)A prediction method of the circulation status of the air traffic network and its segments and its correponding model.The prediction model of the overall traffic circulation status is established using RBF neural network chaotic temporal sequence prediction,and the prediction model of the traffic segments circulation status is established using BP,RBF,GRNN neural network theory.Based on the flight trajectory data of the air traffic network,a prediction method of the traffic circulation status is presented,and proved efficient to provide decision-making support for the optimization and control of air route network traffic circulation.The thesis carries out a profound and systematic research on the perception of air route network traffic circulation.Via acquiring convenient,flight trajectory data that reflects the real-time and accurate air route network traffic circulation status,the study presents methods to evaluate,identify and predict the traffic circulation status of the aire route network and its segments,and establishes their corresponding models.A technical proposal on the perception of traffic circulation status is formed on the basis of flight trajectory data to provide theoretical and technical support for research on a smart perception system of air route network traffic status.