Research On Collaborative Air Traffic Flow Management Based Upon Multi-agent Modeling And Simulation

Collaborative Air Traffic Flow Management (CTFM) is a new concept which seeks to solve thepotential air traffic management problem in the future. The research focus on giving moreopportunities to airspace users to express their preferences while maintains Airlines safety and equity.It is supposed that this will lead to decreased workload for Traffic Management Unit, reduced delaycost, increased airline satisfaction and more efficient air traffic flow management.In this thesis, an agent-based modeling and simulation approach is used to improve thecollaborative air traffic flow management techniques. After analyzing the air traffic behaviors, TrafficManagement Units rules and Airlines Operations Center objectives etc., different classes of agentmodels with corresponding stochastic and dynamic simulation environment are established, andrelative simulation experiments are then conducted to evaluate the performance of the overall systemwhen the air traffic flow management in collaboration with the airlines takes place. The main contentof this dissertation is as follows:(1) Introduce the research fundamental systematically. The operation concept of CollaborativeAir Traffic Flow Management with its origin, content, research objectives and development aredescribed first. After analyzing the detailed air transportation problem in practical, the four phases ofthe ATFM process in the CATFM Concept of Operations are explained. Finally, the paper presents theagent-based model with it communication method and the stochastic simulation modeling theory withdetail.(2) Research has being done to multi-agent based collaborative flights planning problem. First, amulti-level collaborative flight planning procedure is established by means of collaboration betweendifferent air traffic management units and Airlines operations centers. Then, the agent-based modelingmethod is used to construct the core modeling and reasoning modules of air route traffic managementunits. Taking the summer and fall flight schedule of China in2008as data source, the potentialplanning problem of three main hub airports (ZBAA, ZSSS and ZGGG) in China are studied and themulti-objectives flight schedules are generated finally.(3) Research has being done to multi-agent based collaborative air routes management problem.First, an interactive collaborative air routes management procedure is established by means ofcollaboration between air routes traffic management units and Airlines operations centers. Then, theagent-based modeling method is used to construct the core operateing and reasoning modules of airroute traffic management units. After analyzing the radar data from Central and Southern China in 2009, the ARENA tool is used here to conduct a discrete event simulation on the ZGGG-ZBAA routefor validation finally.(4) Research has being done to multi-agent based distributed collaborative aircraft operation.First, an adaptive collaborative aircraft performance procedure is established by means ofcollaboration between adjacnet aircrafts. Then, the agent-based modeling method is used to constructthe core kinetic and reasoning modules of aircrafts by taking use of the point mass model andProportional-derivative Controller. Taking the aircrafts operation performance parameters from theUser Manual for the Base of Aircraft Data published by EUROCONTROL, a combined discretecontinuous simulation of aircrafts flying in a self-separation corridor is conducted to check theperfomances of aircrafts finally.In the end, the thesis summarizes the achievements of the collaborative air traffic flowmanagement based upon multi-agent modeling and simulation, discusses the limitation of thisresearch and makes some suggestions for the future work.