On Dynamic Traffic Information and Patterns of Cooperative Behaviours: Economic Analysis of Information Policies based on a Bi-Layer, Stochastic Traffic Equilibrium Model

An analytical model is developed for a two link network subject to congestion both recurrent and incidental on link travel times, dynamic traffic information (DTI), and two classes of users either informed or not to choose their route. Closed-form formulas are provided for the main model variables, including average link flow and trip cost by user class, as well as marginal external cost by user class, for each of the two basic patterns of cooperation among the network users: User Equilibrium (UE) and System Optimum (SO). Based on parametric analysis, it is shown that under UE the informed users benefit from dynamic information in a decreasing way with respect to demand volume and equipment rate, whereas the uninformed users get benefit only if the link average times are unequal. The potential benefits of system optimization include several terms, of which the balance depends on the system state. Their split between the two classes of users depends on: equal or unequal average link social time; demand volume and equipment rate; and a network configuration parameter that synthesizes the free flow time and capacity features of the network routes. To harvest the benefits of system optimization, while maintaining user acceptability, a pattern of Informed Cooperation is put forward for cost minimization limited to the set of informed users: its effectiveness is demonstrated by numerical simulation