Event-based short-term traffic flow prediction model

The problem of predicting traffic flow for the purpose of real-time traffic-adaptive signal control in an urban street network is explored. A prediction model is described that combines data from traditional vehicle loop detectors and known relationships from traffic flow theory. The model is demonstrated using a microscopic traffic simulation model. Results of the simulation demonstrate that the model can provide the information required to develop truly proactive real-time traffic-adaptive signal control.