Predicting Human Mobility Fluxes Between Regions with Different Features in Urban

Understanding human mobility patterns in urban area is critical to building a better city. Previous researches on human mobility mainly focus on the population distribution and geography, and have revealed a series of statistical law. However, citizens' mobility patterns are not only constrained by trip cost, but also deeply affected by trip purposes. In order to capture both trip cost and trip purpose, we develop a model that combines intervening opportunity theory and transition probability between regions with different features to predict trip fluxes between regions. Results of our model are in good agreements with Shanghai metro data on passenger flows of destination stations, trip time distribution, trip fluxes and its distribution. We also quantificationally show the importance of considering trip purposes and regions features. Our model has a potential guiding significance for predicting the popularity and the changes in trip fluxes when planning new urban districts.