关键词: 交通运输工程；城市交通；建成环境；非线性；随机森林；公交使用

Abstract: Guiding car users to use bus can help alleviate urban traffic congestion and promote energy conservation and emission reduction in the transportation field. Using travel survey data for car users in Kunming City, a random forest model was employed to explore the impact of built environment, bus service quality and individual socio-economic attributes on the use of bus by car users. Research indicates that the built environment is the most important factor influencing the use of bus by car users, with a relative importance sum of 53.7%. Road network density, land use mix, residential location, and the number of bus stops are closely related to the use of bus by car users. Individual socio-economic attributes have a relative importance sum of 24.9%, with the family income variable having the highest relative importance (5.2%). The sum of the relative importance of bus service quality is 21.4%, which is convenience (5.8%), comfort (5.6%), reliability (5.2%) and safety (4.8%), respectively. The impact of the built environment on the use of bus by car users is nonlinear, and there are differences in the nonlinear effects between car users with different income levels. The research results contribute to a better understanding of the bus usage behavior by car users, thus the targeted strategies will be adopted to promote their use of buses.

Key words: traffic and transportation engineering; urban traffic; built environment; nonlinear; random forest; bus utilization