城市大规模交通网络低效路段组合定位及分析
Locating and researching of inefficient road clusters in a large-scale transportation network

基于布雷斯悖论现象,根据城市交通网络中路段或路段组合对路网通行效率的影响,可将其划分为低效的或必要的,合理关闭路网中的低效路段及路段组合将减少所有出行者的出行时间成本,必要路段及路段组合的关闭将增加出行者额外的出行延误。采用改进的自适应遗传算法以有效地定位大规模真实交通网络中的低效路段组合,结合美国旧金山市大规模实际交通网络地理信息系统(GIS)数据以及该区居民通勤出行数据进行实证分析。研究结果证实低效路段和必要路段的组合可能是低效的。分析表明,低效路段组合的低效程度|ΔTS|随着所含路段数量增加呈现先增后减的变化趋势,含路段数相对少的低效路段组合对路网效率产生更大的影响也更普遍。
In this paper, road and cluster of roads in transportation network were classified as inefficient or necessary according to their effect on the efficiency of the road network, based on Braess''''s paradox. Reasonable closure of inefficient links can reduce travel costs considerably, while the failure of necessary links would result in extra travel delays. We modified the adaptive genetic algorithm to pinpoint inefficient road clusters in a real large-scale traffic network. Actual transportation network data from geographical information systems (GIS) and daily commuting origin destination matrix in the San Francisco were used. Our empirical results show that inefficient road clusters include not only inefficient road but also necessary road. Meanwhile, the degree of inefficiency |ΔTS| increases first and then decreases as the number of roads in the road clusters increase. Inefficient road clusters contain less road with large |ΔTS|, which is much more common as well