Investigating the influence of herd effect on the logit stochastic user equilibrium problem

In traditional traffic equilibrium models, travelers typically rely on link flow information to estimate link/route costs. However, an often-overlooked aspect in prior research is that travelers can easily acquire information about the route choices of other travelers within the same OD pair. The presentation of route flow information, combined with the influence of travelers’ herd behavior, results in novel and intriguing conclusions distinct from conventional equilibrium models. Specifically, this study demonstrates that by considering the influence of the herd effect in route choice within the stochastic user equilibrium model, it is possible to address three critical problems that remain inadequately resolved in the literature: (1) The stability of stochastic user equilibrium (SUE) in the presence of route flow information. (2) The existence of meaningful link tolls capable of steering the SUE flow pattern toward system optimum (SO). (3) The design of a trial-and-error congestion pricing scheme with disequilibrium observed network flow patterns. To tackle these problems, we present a logit SUE flow evolution process with herd effect and propose a corresponding trial-and-error congestion pricing scheme. Rigorous proofs of related convergence theorems will be provided. The findings of this study support the assertion that “herd effect can offset travelers’ perception errors”, carrying significant policy implications for leveraging herd effect in the design of navigation software and congestion pricing strategies.