城际集装箱交流枢纽的识别及其物流特征——以中国铁路运输为例

集装箱流反映了城市间的经济联系，枢纽城市在区域经济网络中具有重要地位，主导着物流网络的运行。依据图论原理，构筑了枢纽城市的识别模型，融合了城市交流的直接联系和间接联系，对集装箱枢纽进行识别，发现中国形成了21 个集装箱枢纽，以沿海港口城市为主，枢纽牵引的集装箱交流体系显示“区域-门户”空间结构明显。然后分析了枢纽集装箱流的距离衰减规律，采用Pareto 和Newling 模型进行拟合，确定了主要衰减参数，Pareto 模型的衰减系数b 显示其数值越高，集装箱牵系的城市影响越限于近距离，Newling 模型的递增系数b1为正值说明枢纽附近存在影响塌陷区，最高影响不邻近枢纽。同时依据有向图的矢量属性揭示了枢纽集装箱流的集散规律，发现枢纽城市的辐射力和吸引力呈明显的非均衡性，邻近地区形成影响塌陷区，一定距离范围内形成主影响区，各枢纽的本底决定了其主影响区明显不同，虽然各城市的辐射峰值和吸引峰值不同，但其集聚和扩散通道往往一致。

Identification of Inter-urban Container Transport Hubs and Their Spatial Characteristics:A Case Study of Railway Transportation in China

Container traffic flows reflect the economic linkages inter-cities. Hubs play an important role in regional economic network, which reversely dominates the development of logistics. Based on the graph theory, this paper designed a model, which considered both the direct and indirect economic linkages inter-cities, to identify the hubs. As a result, 21 container hubs were found in China, and most of them are seaport cities. Moreover, the container traffic flow system around the hub shows an obvious region-gateway spatial structure. Then, this paper analyzed the distance decay rule of main container hubs such as Guangzhou, Shanghai and Beijing using Pareto and Newling model respectively, and defined main decay parameters of the models. The decay coefficient b in Pareto model shows that if its value is higher, the container hub will be more limited in a short distance. The increment coefficient b1 in Newling model is positive, which demonstrates that there is a low value area near the hub. Therefore, the area which is influenced most by the container hub is really not close to it. Meanwhile, this paper explored the spatial agglomeration and diffusion rules of container transport based on the O-D flow matrix. Hubs' radiation and attraction capability shows obvious spatial imbalance. Adjacent areas have lower values, and the main influenced areas are found in a certain distance away from the hubs. Although the largest radiation value and attraction value of each hub is different from the others, their agglomeration and diffusion corridors are almost in accordance with each other.