Simulation of Flows in a Stochastic Road Network Based on Sensitivity Analysis

In order to evaluate the uncertainty of system performance in a stochastic road network,a sensitivity analysis based simulation approach to obtain the random link flows is presented.To reflect the stochastic characteristics of travelers' route choice behaviors under uncertain environment,the route-choice dispersion parameter is assumed to be a positive random variable.For any realization of the dispersion parameter variable,the logit-based stochastic user equilibrium(SUE) is employed to obtain the corresponding link flows.So the distribution characteristics of stochastic link flows can be obtained.In order to obviate the need for resolving the SUE problem,the derivatives of SUE link flows with respect to the dispersion parameter are formulated and used to estimate the SUE link flows.A key requirement for computing the derivatives is that the link use probabilities can be computed efficiently.The consistency problem of route sets when calculating joint use probabilities with Dial's stochastic assignment algorithm is analyzed and solved.The feasibility of the proposed approach is illustrated with a numerical example.