Feedback Analysis of Interaction between Urban Densities and Travel Mode Split

A feedback simulation model based on radial basis function neural networks is newly developed in this research to analyse the interaction between urban densities and travel mode split. The changes of populations, road mileages, travel mode split, and so on of the enlarging urbanized areas of different cities in China are studied for the trainings of the radial basis function neural networks constituting the proposed feedback model. Furthermore, the effect of different development policies for Beijing on distinct indicators of the urban density and trip shares of various travel modes is also evaluated by the newly developed model. It is found that stopping the quick urban sprawl of Beijing is the most important for the sustainable development of its urban transport. It is confirmed that the newly developed model is able to rationally explain the interactive correlation between urban densities and travel mode split of a city for its different development plans.