Stochastic/Dynamic User-Optimal Route Choice Model

Abstract

The stochastic dynamic user-optimal route choice model assumes that the perceived travel times are incomplete and/or imprecise. To illustrate the perceived travel times, an error term is often hypothesized to accompany the actual travel times, and many probability distributions have been applied to represent the real situation. For a specified error term of the perceived travel times, we assume all drivers make their route choice decisions based on their perceptions of O-D travel times. A notable example is the Gumbel distribution which yields the logit model. The resulting traffic flows are expected to be more dispersed over parallel routes for the stochastic dynamic route choice model than for its deterministic counterpart.