A large-scale multi-objective flights conflict avoidance approach supporting 4D trajectory operation  

作　　者：Xiangmin GUAN Xuejun ZHANG Renli LV Jun CHEN Weiszer MICHAL 

机构地区：[1]Department of General Aviation, Civil Aviation Management Institute of China [2]School of Electronic and Information Engineering, Beihang University [3]School of Engineering, University of Lincoln

出　　处：《Science China(Information Sciences)》2017年第11期194-206,共13页中国科学（信息科学）（英文版）

基　　金：supported by National Natural Science Foundation of China (Grant Nos. U1533119, U1433203);Foundation for Innovative Research Groups of the National Natural Science Foundation of China (Grant No. 61221061).

摘　　要：Recently, the long-term conflict avoidance approaches based on large-scale flights scheduling have attracted much attention due to their ability to provide solutions from a global point of view. However, current approaches which focus only on a single objective with the aim of minimizing the total delay and the number of conflicts, cannot provide controllers with variety of optional solutions, representing different tradeoffs. Fur- thermore, the flight track error is often overlooked in the current research. Therefore, in order to make the model more realistic, in this paper, we formulate the long-term conflict avoidance problem as a multi-objective optimization problem, which minimizes the total delay and reduces the number of conflicts simultaneously. As a complex air route network needs to accommodate thousands of flights, the problem is a large-scale combinato- rim optimization problem with tightly coupled variables, which make the problem difficult to deal with. Hence, in order to further improve the search capability of the solution algorithm, a cooperative co-evolution （CC） algorithm is also introduced to divide the complex problem into several low dimensional sub-problems which are easier to solve. Moreover, a dynamic grouping strategy based on the conflict detection is proposed to improve the optimization efficiency and to avoid premature convergence. The well-known multi-objective evolutionary algorithm based on decomposition （MOEA/D） is then employed to tackle each sub-problem. Computational results using real traffic data from the Chinese air route network demonstrate that the proposed approach ob- tained better non-dominated solutions in a more effective manner than the existing approaches, including the multi-objective genetic algorithm （MOGA）, NSGAII, and MOEA/D. The results also show that our approach provided satisfactory solutions for controllers from a practical point of view.Recently, the long-term conflict avoidance approaches based on large-scale flights scheduling have attracted much attention due to their ability to provide solutions from a global point of view. However, current approaches which focus only on a single objective with the aim of minimizing the total delay and the number of conflicts, cannot provide controllers with variety of optional solutions, representing different tradeoffs. Fur- thermore, the flight track error is often overlooked in the current research. Therefore, in order to make the model more realistic, in this paper, we formulate the long-term conflict avoidance problem as a multi-objective optimization problem, which minimizes the total delay and reduces the number of conflicts simultaneously. As a complex air route network needs to accommodate thousands of flights, the problem is a large-scale combinato- rim optimization problem with tightly coupled variables, which make the problem difficult to deal with. Hence, in order to further improve the search capability of the solution algorithm, a cooperative co-evolution （CC） algorithm is also introduced to divide the complex problem into several low dimensional sub-problems which are easier to solve. Moreover, a dynamic grouping strategy based on the conflict detection is proposed to improve the optimization efficiency and to avoid premature convergence. The well-known multi-objective evolutionary algorithm based on decomposition （MOEA/D） is then employed to tackle each sub-problem. Computational results using real traffic data from the Chinese air route network demonstrate that the proposed approach ob- tained better non-dominated solutions in a more effective manner than the existing approaches, including the multi-objective genetic algorithm （MOGA）, NSGAII, and MOEA/D. The results also show that our approach provided satisfactory solutions for controllers from a practical point of view.

关 键 词：air traffic management conflict avoidance combinatorial optimization MULTI-OBJECTIVE coopera-tive co-evolution 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程] V355[自动化与计算机技术—控制科学与工程]