粤港澳大湾区内地城市群路网韧性的贝叶斯网络评估

Bayesian network evaluation of road network resilience in the Guangdong-Hong Kong-Macao Greater Bay Area city cluster

随着我国城市化进程的推进,城市路网的韧性问题愈发凸显。构建了一种基于贝叶斯网络模型的城市路网韧性评价模型,并以粤港澳大湾区内地城市群为例,评估了2000—2019年该地区9个城市的路网韧性。基于韧性功能函数从容灾性、稳定性、重构性、恢复性4个方面建立功能层,选取稳定性、可变性、可维护性、可靠性、服务性、安全性、鲁棒性、可修性、适应性9个因素建立性能层,选取客运量、公共汽车数量、道路面积率、出租车数量、交通事故死亡人数、公共交通投资、每万人公共汽车数、道路运输从业人数、人均道路面积9个因素建立因素层。研究结果表明:路网韧性指数最高的为深圳(60%),而最低的为中山(43%);广州、深圳、珠海、东莞的路网韧性一直稳定增长,其余城市的路网韧性浮动较小;路网韧性并非总是随着城市GDP的增长而提高,还与人均GDP有关;敏感性分析表明,路网重构性、路网可靠性、交通事故死亡人数分别为各自所在层中对路网韧性影响最大的因素;加强中心城市与其他城市之间的联系可以提高城市路网韧性。研究成果可为粤港澳大湾区城市群交通路网功能完善与韧性提升提供技术参考。

With the further increase of the urbanization rate in China,the resilience of urban road networks has attracted increasing attention.In this paper,a resilience evaluation model for the urban road network is developed based on the Bayesian network model.Then,taking the city cluster in the mainland of the Guangdong-Hong Kong-Macao Greater Bay Area as an example,the road network resilience of nine cities in the mainland part of the Guangdong-Hong Kong-Macao Greater Bay Area during the period 2000-2019 is evaluated.According to the resilience function,the functional layer is constructed based on the four indicators as resilience,stability,reconfigurability,and recovery,the nine factors as stability,variability,maintainability,reliability,serviceability,safety,robustness,repairability,and adaptability are selected to establish the performance layer,and the nine factors as passenger volume,number of buses,road area rate,number of cabs,number of traffic fatalities,public transportation investment,number of buses per 10000 people,number of road transportation employees,and road area per capita are selected to build the factor layer.The results show that Shenzhen‘s road network resilience index is the highest at 60%,while Zhongshan‘s is the lowest at 43%.The road network resilience of Guangzhou,Shenzhen,Zhuhai,and Dongguan has been growing steadily during the period 2000-2019,while that of Foshan,Huizhou,Zhongshan,Jiangmen,and Zhaoqing fluctuates slightly.The urban road network resilience does not always increase with the GDP growth but is also related to per capita GDP.The sensitivity analysis for the elements of function layer,performance layer and factor layer in the Bayesian network model shows that the road network reconfigurability,road network reliability,and number of traffic fatalities are the most influential factors on road network resilience in their respective layers,and strengthening the links between the central city and other cities can improve the resilience of urban road networks.The research results of this paper can provide some technical references for the improvement and resilience growth of the transportation network function in the Guangdong-Hong Kong-Macao Greater Bay Area city cluster.