Abstract: Network heterogeneity cannot be ignored for traffic congestion analysis, because it directly associates with network structure. This paper investigates the relationship between network heterogeneity index and network size, and analyzes the influence of heterogeneity index on traffic congestion. For this aim, the paper presents the definition of network heterogeneity index and establishes the traffic flow model of heterogeneous structure. The results show that the heterogeneity decreases with the network size. When network size is a fixed value and the heterogeneity index is at subcritical region, traffic congestion increases with heterogeneity index. Furthermore, it can also be found that when the heterogeneity index reaches the critical value, the average number of accumulated vehicles reaches its maximum and congestion is significantly serious. Therefore, those networks, in which heterogeneity index is smaller than or larger than the critical value, have more tolerant to congestion. Adjusting the heterogeneity index can improve the congestion of traffic network to some extent.

Key words: highway transportation, congestion, heterogeneity index, traffic network, complex networks

摘要： 网络异质性与网络结构有着直接的关联，是分析交通拥塞不可忽略的因素.本文研究网络异质性与网络规模的关系，并分析网络异质性指标对交通拥塞的影响.为此，给出网络异质性指标的定义，并建立基于异质结构的交通流模型.结果表明，网络异质性随着网络规模的增加而减小.当网络规模一定且在异质性小于临界值时，拥塞随着网络异质性的变大而增加.进一步研究发现，在异质性到达临界值时，路段平均输出车辆数最小而网络中的累积车辆到达最大，此时拥塞最为严峻.因此，具有小于和高于异质性指标临界值的网络具有更好的抗拥塞能力，调节网络的异质性指标可以改善交通网络的拥堵.

关键词: 公路运输, 拥塞, 异质性指标, 交通网络, 复杂网络