Understanding personal travel behavior can provide a theoretical basis for urban and transportation planning. The majority of the studies are based on aggregated individual data, and there are few microscopic studies at the household level. The paper takes the household as the research object, and uses the improved activity space as the behavior index, and based on the 2016 Kunming data clustering, five typical household spatial Clusters are obtained. The study found that there are differences in commuting distance, travel frequency and transportation mode among different households. Through multiple logit models, the household economic attributes and built environment are affected on household activity space. Increasing the density of entertainment facilities in the male active area is conducive to narrowing with higher activity spatial diffusion. Increasing the density of retail commercial facilities in the active area can reduce the distance traveled by female.

CICTP