枢纽机场终端区航班协同调度模型及快速智能优化算法研究

The contradiction between the rapid growth demand of air traffic and the limited airspace resource around hub airports is becoming increasingly prominent. So how to improve the automation level for aircraft scheduling of airport terminal areas has become a problem to be solved urgently. Aircraft scheduling of airport terminal area is a NP-hard problem with multi-constraint and collaboration characteristic. Its core is the collaborative aircraft scheduling model and the fast sorting algorithm. The idea of this project is to set up arrival and departure aircraft scheduling models respectively by analyzing the airspace, the airport layout and other constraints, and study the relationship between two scheduling models, then design especial data structures to share resource and study the mechanisms of constraints transmission, finally establish the collaborative aircraft scheduling model of airport terminal area. Combined with the suppression and dynamic adjustment of the immune network, and the application of domain knowledge of aircraft scheduling, the fast immune particle swarm optimization algorithm based on domain knowledge will be researched, and also implement the optimization solution to the collaborative aircraft scheduling model. In the dynamic process of the collaborative aircraft scheduling,we will study methods of resource request time prediction, dynamic resource allocation and conflict detection to achieve a simulation system of dynamic aircraft collaborative scheduling. The innovation of this project is to research on the collaborative aircraft scheduling model of airport terminal area and the fast immune particle swarm optimization algorithm based on domain knowledge.

急剧增长的空中交通需求与有限的枢纽机场空域资源间的矛盾日益尖锐，如何提高机场终端区航班调度自动化水平成为一个亟待解决的难题。终端区航班调度问题是一类具有多约束协同特性的NP难题,其核心是航班协同调度模型和快速排序算法。本课题的基本思想是在系统研究空域结构、机场布局等多种约束的基础上，分别建立进场调度和离场调度数学模型，并研究两种调度之间的制约关系，通过设计共享资源描述数据结构、分配机制和约束传递机制，建立终端区航班协同调度模型；结合免疫网络的抑制和动态调节机制、航班调度领域知识的合理应用，研究基于领域知识启发的免疫粒子群快速优化算法，并实现对航班协同调度模型的优化求解；结合航班协同调度的动态运行过程，研究资源请求时间预测、动态资源分配和冲突检测方法，设计实现动态航班协同调度仿真系统。创新之处是提出开展终端区航班协同调度模型和基于领域知识启发的免疫粒子群快速优化算法的研究。

急剧增长的空中交通需求与有限的枢纽机场空域资源间的矛盾日益尖锐，课题组紧密围绕如何提高机场终端区航班调度的自动化水平这一难题做了系统、深入、持续的研究。首先对现行航班航路调度、进近调度、塔台调度和场面调度方式及各种约束条件进行了系统的研究，分别建立了进离场调度的数学模型，并着重研究了两种调度之间的相互制约关系，通过设计共享资源描述数据结构、分配机制和约束传递机制，建立终端区航班协同调度模型；深入民航一线单位收集、整理了大量的雷达和QAR（Quick Access Record）历史数据集，运用统计分析、模糊关联规则、模糊分类方法挖掘和提取航班调度的经验知识，提出了高效的基于领域知识启发的经验粒子群优化算法和其他智能仿生算法，并运用这些算法实现了对航班协同调度模型的优化求解；结合航班协同调度的动态运行过程，提出基于资源请求时间预测的动态资源分配和冲突检测方法，设计实现了动态航班协同调度仿真系统。创新之处是针对问题的多约束协同特性，建立了终端区航班协同调度模型；针对问题的实时特性，提出了高效的基于领域知识启发的经验粒子群优化算法。