Abstract: Long-time commuting has become a common issue faced by urban residents in China, seriously affecting the quality of life and commuting efficiency of residents. Taking the built-up area of Kunming as a case study, the nonlinear relationships and threshold effects between the built environment and the long-time commuting of urban residents were explored by the random forest model. The research results show that the sum of the relative importance of the built environment to residents long-time commuting is 80.8%, far greater than individual socio-economic attributes. The self-sufficiency index has the greatest and negative impact on long-time commuting, followed by population density, road network density and employment density. Furthermore, there are nonlinear relationships between the built environment and long-time commuting of residents. Population density, employment density, road network density, job-to-residence ratio, and bus stop density all exhibit clear threshold effects on long-time commuting.

Key words: traffic and transportation engineering; transportation planning and management; built environment; long-time commuting; nonlinearity; random forest

摘要： 长时通勤已成为我国城市居民面临的普遍问题，这严重影响了居民的生活质量和通勤效率。以昆明市建成区为例，采用随机森林模型探究了建成环境与城市居民长时通勤的非线性关系与阈值效应。研究结果表明：建成环境对居民长时通勤的相对重要度之和为80.8%，远大于个人社会经济属性；自足性指数对长时通勤的影响最大且为负相关，其次为人口密度、路网密度、就业岗位密度；建成环境与居民长时通勤之间存在非线性关系，人口密度、就业岗位密度、路网密度、职住比、公交站点密度对长时通勤均表现出明显的阈值效应。

关键词: 交通运输工程；运输规划与管理；建成环境；长时通勤；非线性；随机森林