Modeling roadway network traffic using a generalized radiation model
Our work was motivated by the need to understand, model and predict transport processes in complex networks and in particular in roadway transportation infrastructure networks. We present a method for modeling traffic flows on roadway networks by introducing a generalized version of the radiation model as traffic law, coupled with an efficient, capacity aware range-limited network flow distribution algorithm. The radiation model is extended via a general travel cost function allowing us to compare different cost criteria such as travel distance vs travel time in their ability to predict real traffic flows. Using census data and US highway network traffic data from MIT’s GIS database for validation, we show that the travel time cost based model captures the distribution of the annual average daily traffic and attains a high Pearson correlation coefficient (0.75) when compared to real traffic flows.