Stochastic Equilibrium Model of Peak Period Traffic Congestion

This paper addresses the problem of peak period traffic congestion. It considers the queues and delays at a single point of insufficient capacity. A model is developed to predict the pattern of traffic volumes and travel times during a peak period. It consists of two basic elements: a deterministic queue and a random utility departure time choice. The utility represents the tradeoff that exists whenever congestion occurs at the desired departure time. A trip-maker can then shift his/her trip forward or backward in time to avoid a long delay. The properties of the equilibrium solution of the model are investigated analytically. It is shown that there cannot be more than one congestion period and that the equilibrium solution is unique.