Freeway travel time systems are becoming more prevalent throughout the world, and while they represent a substantial improvement in the provision of information to motorists, their credibility and effectiveness is closely linked to their level of accuracy. This paper presents a new macroscopic algorithm for predicting freeway travel times. The proposed algorithm has the ability to be used in any traffic situation between any two points. It also has the ability to account for the relative speeds and densities of vehicles within the traffic stream. Calibration and field testing of the new algorithm has indicated that substantial improvements in travel time prediction can be achieved when compared to an existing system currently operational in Melbourne, Australia. (A*)
Abstract