Abstract: Promoting the change of urban residents’ short-distance travel mode from motor vehicles to slow traffic is of vital importance to alleviate traffic congestion and environmental pollution caused by motorized travel. Based on the travel survey data of Kunming residents, travel attributes and built environment data, a geographically weighted logistic regression model was used to explore the influence of travelers’ socioeconomic-demographic attributes, travel attributes and the built environment on the choice of cars and slow traffic modes in short-distance travel. The results show that the geographically weighted logistic regression model fits better than the global regression logistic model does. Intersection density, road network density, population density, total POI density, retirees and commuter trips have a negative effect on car use in short-distance travel in general and a positive effect on distance to CBD in general. But the effects of these variables are significantly spatially heterogeneous, showing opposite effects in different regions. In addition, men, self-employed individuals, groups under the age of 30, college graduates and above, and groups with higher household income are more likely to use cars for short-distance trips, while shopping and entertainment trips are more likely to use slow traffic.

Key words: traffic engineering; short distance travel; geographically weighted logistic regression; travel mode choices; spatial heterogeneity; built environment

摘要： 促进城市居民短距离出行方式由机动车向慢行交通转移，对缓解机动化出行造成的交通拥堵和环境污染至关重要。基于昆明市居民出行调查数据和建成环境数据，利用地理加权逻辑回归模型探究了出行者社会经济-人口属性、出行属性与建成环境对短距离出行中小汽车和慢行交通方式选择的影响。研究表明：地理加权逻辑回归模型拟合结果优于全局回归逻辑模型。交叉口密度、道路网密度、人口密度、总POI密度、离退休人员和通勤出行在总体上对短距离出行中的小汽车使用起到负向影响，到CBD距离在总体上起到正向影响，但这些变量的影响具有显著的空间异质性，在不同区域表现出相反的作用。此外，男性、个体经营者、30岁以下、大专及以上、家庭收入越高的群体更倾向于在短距离出行中使用小汽车，而购物和娱乐出行更倾向采用慢行交通方式。

关键词: 交通工程；短距离出行；地理加权逻辑回归；出行方式选择；空间异质性；建成环境