【作者】 卫东选；

【导师】 刘长有；

【作者基本信息】 南京航空航天大学 ， 交通运输规划与管理， 2010， 博士

【摘要】 停机位是民用机场中的重要运行设施和资源,为每个进离港的航班选择并分配合理的停机位是机场运行管理中的一项核心任务。现有的停机位分配问题研究主要基于运行效率等方面,对于停机位分配时飞机的运行安全性考虑较少。本文以避免飞机推出冲突为切入点,运用安全管理中防患于未然的管理理念,在概念创新和原有只考虑效率研究的基础上,系统研究了兼顾运行安全和运行效率的停机位分配优化问题。本论文详细讨论了现有停机位分配相关问题的研究现状,分析了现有研究中的不足之处,指出实现系统运行安全和运行效率双赢是本文研究停机位分配问题的新思维。文章对机场停机位分配问题的研究对象进行了描述,给出了建模需要考虑的约束条件和现有主要优化目标的数学描述,并对影响问题求解的因素和问题求解方法进行了讨论。在对飞机机坪运行冲突进行分析的基础上,研究了机位资源充足时考虑运行安全的停机位分配问题。提出以“关口前移”的主动避免方法,将限制推出冲突的条件作为安全性约束,建立了推出冲突避免的单目标停机位分配问题新模型。在问题规模较小时,采用整数规划对其进行精确求解;在问题规模增大时,考虑到变量和约束条件数量快速增长带来的组合爆炸问题,设计粗粒度并行遗传算法进行求解。在停机位资源充足的条件下,该模型可达到避免冲突且尽量提高运行效率的目的。考虑到繁忙机场高峰时段经常出现的停机位资源暂时短缺现象,以及严格约束可能导致的不可解性问题,提出以解除冲突导致的相关旅客等待时间来评价分配方式的运行安全性,建立机坪冲突影响最小化的过约束停机位分配问题新模型,并采用粗粒度并行遗传算法进行求解,得到的优化分配方式可有效降低飞机的邻接机位冲突、机坪滑行冲突,以及停机位资源短缺时分配到同一机位的连续航班对机位利用的冲突等不安全因素的影响。对机位资源充足时的推出冲突避免多目标机位分配问题进行了研究。采用机位空闲时间段的模糊隶属度代替确定性的0-1关系来描述航班-机位之间的匹配程度,并设计基于航班-机位模糊隶属度的调节函数,将分配方式鲁棒性和旅客行走距离最小两个重要优化目标转换为单目标,采用粗粒度并行遗传算法进行优化,在避免推出冲突条件下达到两个相矛盾目标之间的妥协优化。考虑运行安全和运行效率等目标,研究了过约束的多目标停机位分配问题,优化目标包括基于运行安全的相关旅客等待时间最小,以及旅客模糊行走距离最小、行李模糊搬运距离最小。设计基于Pareto排序的多目标粗粒度并行遗传算法对其进行优化,优化结果可获得一组在Pareto前沿分布均匀的Pareto最优解集合,供决策者选择。此外,对现有停机位分配问题的主要研究方向进行了分析和改进。分析和设计了预分配方式鲁棒性目标的优化算法;对旅客行走距离最小目标的优化特点进行了分析;对机位资源受限时考虑机型匹配的航班靠桥率最大化优化算法进行了研究;对实时运行中部分航班时刻发生改变的停机位实时再分配问题进行了分析和建模,并采用启发式算法和禁忌搜索算法进行优化。更多还原

【Abstract】 Gates are important facilities and key resources in civil airport. Selecting and assigning available gate for each arriving and departing flight is a key activity in airport operations. The existing researches on gate assignment problem are mainly based on operational efficiency, while aircraft safety about the gate assignments almost could not be considered in details. Based on conceptual innovation and existing researches on efficiency, this paper will study the airport gate assignment problem (AGAP) which considering both operational safety and efficiency by using the philosophy of“nip in the bud”in safety management.This thesis discusses AGAP in detail and points out deficiencies of existing researches. New idea of this paper for studying AGAP is achieving the win-win situation of operational safety and efficiency. Airport gates and flights are described. And then constraints and existing main optimal objectives are formulated, and factors affecting problem solving and approaches are discussed. Based on the analysis of the operational process of aircrafts on the apron, gate assignment problem considering operational safety is studied when gates are adequate at first. The proactive approach of“avoid conflict in advance”is adopted by regarding restricted conditions of potential power-in vs. push-out conflict as safety constraints, and then the model with conflict-avoidance is proposed. A mathematical programming technique is used to achieve optimal resolution when the problem scale is small. While the scale is much larger, coarse-grain parallel genetic algorithms(CPGA) is designed to get satisfactory solutions by taking into account the rapid growth of the number of variables and constraints which contribute to combinatorial explosion. The ultimate purpose is enhancing the operational efficiency with avoiding conflict.Given to the heavy traffic of airports, temporary gate shortages are common phenomenon. Considering the algorithmic unsolvability for gate shortages and strict constraints, minimizing waiting time of passengers to relieve conflicts is proposed as optimization objective, and new model based on operational safety of gate assignment problem is presented. CPGA is adopted to solve the model. The optimization of the model can get the safest assignment while reducing unsafe factors such as conflicts of power-in/push-out, taxiing conflicts in apron, and utilizing conflicts for temporary gate shortages.Multi-objective optimization model with conflict-avoidance is studied while gates are adequate. The fuzzy membership grades of gate idle periods are used to describe the matching degree of flight-to-gate instead of the determinate 0-1 relationship. Adjustment function on membership degree is introduced to transfer two objectives into one. CPGA is adopted to get a reasonable trade-off between these two conflict objectives based on avoiding conflict.Over-constrained multi-objective model which considering both operational safety and efficiency is studied. The objevctives inculde minimizing the waiting time of relative passengers based on safety, the fuzzy walking distance of total passengers and the fuzzy baggage transport distance. Multi-objective coarse-grain parallel genetic algorithm is designed to solve the problem. The optimization result can get a set of Pareto solutions which evenly distributed in the Pareto front for airport managers’choice.In addition, some main objectives of existing gate assignment problem are analyzed and discussed. The algorithm for optimizing the robustness of gate assignment is presented. The optimization characteristic of minimizing the total passengers walking distance is analyzed. The algorithm for maximizing the rate of flight-to-bridge while gates are heterogeneous is developed. The gate reassignment problem is analyzed and modeled while flight schedule are changed, and heuristic algorithm and tabu search algorithm is adopted to optimize the model.更多还原