User-Optimized Traffic Assignment

If drivers select routes to minimize their travel times, then the division of traffic among alternative routes of a network is predicted accurately by an algorithm that utilizes realistic increases in travel time with increasing congestion and assigns traffic flow so that travel times on alternative routes carrying traffic are equal and less than travel times on routes not carrying traffic. Such a user-optimizing assignment algorithm is developed from concepts of graph theory. A computer program based on the algorithm was used to generate an example assignment, which is compared to a system-optimized assignment for the same network. Running time for the program was only several times as great as would be required for the most primitive assignment technique. The user-optimizing assignment algorithm should be valuable in conducting experimental investigations of factors affecting route selection and may reduce the uncertainties of highway planning.