مقاله سال ۲۰۱۶

Over the last twenty years, the air cargo industry in China has seen a rapid expansion. In particular, with the rapid expansion of ecommerce markets, the amount of cargo transported by such express companies has increased continuously. For example, the cargo transported by ShunFeng (SF) Express Company by air in 2014 reached 0.83 million tons, increasing 84.4% compared with 2011. However, Chinese air express companies typically only own a few cargo aircraft, and their capacity is far behind the increasing demand. For example, the SF Express Company, one of the largest such companies in China, has established a flight network covering 80% of the main cities; however, the company only has 15 cargo aircraft. Due to this shortage of cargo aircraft, Chinese air express companies have to rent the bellies of passenger aircraft from passenger airline companies and provide transportation services under the hybrid model of “bellies þ cargo aircraft”.

In China, direct passenger flights are currently operated between most main cities. Thus, the flight network of the bellies of passenger aircraft transportation mode presents a “point-to-point” configuration, and cargo aircraft are operated between mega-cities to maximize the loading ratio. In this case, the cargo in cargo aircraft contains cargo traveling directly from one mega city to another mega city and cargo transshipping from the bellies of passenger aircraft. Thus, cargo aircraft and belly transportation utilize a “hub-and-spoke” configuration. As the result, the Chinese air express transportation network is a mixture of “point-to-point” networks and “hub-and-spoke” networks. In this mixed network, due to short of cargo aircrafts, the rented bellies play the main role. This study designs the structure of the mixed air cargo transport network, in which the bellies of passenger aircraft and cargo aircrafts are jointly used. Then we optimize the operation plans for transporting cargos between O-D pairs based on the mixed network. With the calculated results, we further analyze the paths of cargos and discuss the transshipments at some airports.

The remainder of this paper is organized as follows. Next section reviews the relevant literature. In section 3, the problem of this paper is described. Section 4 constructs a bi-level programming model to tackle the problem. In section 5, a genetic algorithm is developed to solve the upper model and the lower model is solved with the algorithm of Frank-Wolfe. Section 6 tests the constructed model with a case study. In section 7, this paper ends with some conclusions of the research.