文章编号： 1672-5328（2020）05-0110-10

姚广铮1, 2，刘小明1，陈艳艳1，崔凯俊2
(1. 北京市交通工程重点实验室，北京工业大学，北京100124；2. 南京市城市与交通规划设计研究院股份有限公 司，江苏南京210002)

摘要： 准确判断城市汽车拥有水平对于汽车产业发展和城市交通规划具有重要意义。利用解释结构 模型，构建影响城市汽车拥有水平的整体框架体系，将15 个影响因素划分为4 个层次。识别了城市 人口规模、城市人口密度等影响汽车拥有水平的根本性因素，以及居民可支配收入水平、公共交通 服务水平等最直接的影响因素。构建城市建成区人口密度、城市人口规模与千人汽车保有量极限值 的回归模型，发现城市建成区人口密度相比于城市人口密度具有更好的解释力，负指数模型比线性 模型有更好的解释力。通过模型推算，中国超大城市、特大城市和大城市远期的千人汽车保有量将 分别处于300 辆·千人-1，350 辆·千人-1和400 辆·千人-1左右的水平，小城市可能达到450 辆·千人-1甚 至更高水平。

关键词： 交通政策；汽车保有量；影响因素；解释结构模型；回归分析；千人汽车保有量

中图分类号： U491

文献标识码：A

Forecasting Maximum Automobile Ownership in Urban Areas

Yao Guangzheng1, 2, Liu Xiaoming1, Chen Yanyan1, Cui Kaijun2
(1.Beijing Key Laboratory of Traffic Engineering, Beijing University of Technology, Beijing 100124, China; 2.Nanjing Institute of City & Transportation Planning Co., Ltd., Nanjing Jiangsu 210002, China)

Abstract： Accurately identifying the level of urban automobile ownership is important for the development of automobile industry and urban transportation planning. Utilizing the cause-effect interpretive structural model (ISM), this paper develops a framework system that explore the relationship between urban automobile ownership and 15 influential factors in four levels. The urban population and population density are the fundamental factors to automobile ownership, and residents' income and public transportation level of service directly affect the automobile ownership. The developed regression model reveals that the population density in urban built-up areas is more explicable than the overall urban population density to automobile ownership, and the negative exponential model works better than the linear model. The modeling results show that the long- term automobile ownership in China's megacities, supercities and large cities will be around 300, 350, and 400 automobiles per thousand people respectively, while the ownership in small cities may reach 450 automobiles per thousand people or even higher levels.

Keywords： transportation policy; automobile ownership; influencing factors; interpretive structural model; regression analysis; automobile ownership per thousand people