Urban spatial structure and travel patterns: Analysis of workday and holiday travel using inhomogeneous Poisson point process models

City land-use features and travel behavior are mutually related and restricted. This research attempts to model the spatial-temporal travel patterns based on spatial point pattern theory. Using a twelve-day private automobile data set collected in Beijing, we systematically investigate the temporal variations of trip-destination distributions, and their association with city spatial structure. The availability of detailed POIs (Points of Interest) data enables us to study the effect of city structure on travel pattern at a refined level. Four types of inhomogeneous Poisson point process models are built to capture the impacts on human mobility posed by spatial covariates. Residual analysis, inhomogeneous K function and leverage diagnostic tools are further adopted to validate the model performance and determine the best fitted model. The validation results indicate that the proposed model reasonably explains the travel patterns in both holiday and workday throughout the city. The inclusion of Cartesian coordinates, population distribution, and city subdivision-category improves the model performance. The empirical results based on the dataset also reveal the differences in impacts on travel patterns posed by underlying city structure between holidays and weekdays as well as between citywide districts. The modeling method and the exploratory spatial–temporal analysis in this study can offer complementary techniques for traffic management and urban planning.