基于多目标进化的航路网络汇聚点设计方法研究

Model and Algorithms for Air Route Network Optimization

航路网络是实现空中交通的物理空间,其结构的科学性与合理性制约着我国民航业及航空运输业的发展。本文把航路网络规划问题简化为一个等价的优化问题——航路网络汇聚点优化布局问题（CWLP问题）。综合考虑航路网络规划的经济性和安全性,同时考虑到航路网络实际布设时的约束,将CWLP问题建模成带约束的双目标优化问题。为了求解该优化问题,提出了基于全面学习蚁群算法的多目标优化算法,该算法具有较强的全局搜索能力,能同时优化多个目标,较好地解决了CWLP问题。最后通过提出的算法与现行航路网络在经济性和安全性方面进行对比,本文算法优化后的方案显示出了较强的优越性,从而证明了本文关于航路网络汇聚点设计方法的有效性。

Air route network（ARN） is the physical space for the air traffic control. The scientific design of the ARN structure is the key to the development of civil aviation and air transport industry. In this paper, the ARN design problem is simplified to an equivalent optimization problem – the crossing waypoint location problem（CWLP）. We integrate both the financial and the safety measurements on the CWLP as well as the real-world constraints on placing waypoints to formularize the CWLP into a constrained bi-objective optimization problem. We proposed a comprehensive learning particle swarm optimization based multi-objective algorithm to tackle the CWLP. This newly proposed algorithm has great global searching ability and is thus very suitable to the CWLP problem. The comparison between the design plan obtained by our algorithms and the existing location plan showed great superiority of our solutions, which demonstrates the validity and efficiency of the proposed air route network design algorithms.