基于可达性指标的航空货运网络布局优化研究

以网络可达性和利用率最大为目标,建立货机班次分配与货物运输路径选择的多目标双层优化模型,针对我国基于审批方式为各航线分配航班的实际情况,通过调整货机班次优化航空货运网络布局.以20个机场组成的航空货运网络为例,设计NSGAII算法求解模型.结果表明:北京、上海、广州、深圳之间的航段将获得1/4的货机班次,他们是航空货运的枢纽机场;当决策者偏好网络可达性时,更多货机班次将被分配到大型机场间的航段上,当决策者偏好网络利用率时,货机班次在航段上的分配较均匀;随着货机班次在航段上由集中趋向分散,网络可达性降低而利用率提高;减少货机班次会降低网络可达性,增加货机班次会降低网络的利用率.

The Optimal Distribution of Air Cargo Transportation Network Based on Accessibility Index

Civil Aviation Administration controls flights distribution on different segments in China. Based on this fact, this paper tends to optimize distribution of air cargo transportation network by adjusting cargo aircraft flights. The purpose is to effectively allocate air cargo transportation resource and provides decision support for macro-control policy. Aiming to simulate the multi-objective decision process of network design problem, a multi-objective bi-level programming model combined with cargo transportation route selection is proposed to maximize network accessibility and utilization. The model is solved by NSGA-II algorithm and the trade- offs between the objectives are reflected by Pareto optimal solutions. Taking 20 airports in China as an example, air cargo transportation network distribution is analyzed. The results show: (1) A quarter of cargo aircraft flights is distributed to the segments between Beijing, Shanghai, Guangzhou and Shenzhen, which are considered as hub airports in the network. (2) When the decision maker prefers to network accessibility, more cargo aircraft flights should be distributed to segments between hub airports. On the contrary, cargo aircraft flights should be evenly distributed to segments if average network utilization is preferred. (3) If cargo aircraft flights distribution tends to disperse on the segments, the network accessibility will decrease, but the network utilization will increase in the optimized schemas. (4) Reducing the total number of cargo aircraft flights will decrease network accessibility. On the other hand, increasing the total number of cargo aircraft flights will decrease network utilization.