Cross Comparison of Spatial Partitioning Methods for an Urban Transportation Network
Cross Comparison of Spatial Partitioning Methods for an Urban Transportation Network
复制标题
城市交通网络空间分区方法的交叉比较
DOI：
10.1007/s13177-019-00209-x
复制
发表时间：
2019
影响因子：
2.1
通讯作者：
Qian Ge and Daisuke Fukuda
中科院分区：
文献类型：
--
作者：
Takao Dantsuji;Shin Hirabayashi;Qian Ge and Daisuke Fukuda
将异质道路网络划分为同质子网络是解决高散射或滞后回线问题的一种方法，该问题可能是宏观基本图的几何学中固有的。本研究进行交叉比较分析的两个良好的研究分区方法-（一）社区检测通过模块化最大化和（二）规范化的割图分区-调查这些方法的适用性。通过一个案例研究，使用真实的交通数据记录了大量的检测器在东京中央商务区，我们发现，这两种方法都工作良好的测试交通网络，但是，不理想的结果可能会得到，如果只有一个瓶颈，在一个子网中，或者如果有一个剧烈的变化，相邻的链接之间的交通状况。
Partitioning a heterogeneous road network into homogenous subnetworks is one way of solving the problem of high scattering or a hysteresis loop that may be inherent in the empirics of a macroscopic fundamental diagram. This study conducts cross-comparison analysis of two well-studied partitioning methods—(i) community detection through modularity maximization and (ii) normalized cut graph partitioning—to investigate the applicability of these methods. Through a case study using real traffic data recorded by an enormous number of detectors in the Tokyo central business district, we found that both methods work well for the test transportation network; however, undesirable results may be obtained if there is only one bottleneck in a subnetwork, or if there is a drastic change in traffic conditions between adjacent links.