Travel time predictions in urban networks.

In-vehicle information, collective and individual dynamic route guidance, congestion management, and incident detection are all systems requiring predictions of short-term link travel times. Hobeika et al. (1993) were among the early researchers who provided a complete approach for predicting travel times in urban networks under incidents based on anticipatory traffic conditions. This paper extends the early work to include travel time predictions under lane(s) closure for construction and maintenance. The algorithm for lane closure relies on the real time dynamic data obtained from sensors and uses a macroscopic traffic input-output model for calculating delays and consequently travel times. The algorithm has been tested and verified using the well-known CORSIM simulation model developed by the Federal Highway Administration. For the covering abstract see ITRD E128239.