Canadian Journal of Civil Engineering · 2011 · 16 citations · 23 references
PlanningTransportation PlanningStochastic ProcessesTransportTransportationMonte Carlo MethodsSimulation MethodsTraffic Flow TheoryTransport ModelingTransport ModellingEnsemble MeanEnsemble VarianceOrigin-destination EstimationTravel Time Variability
Daily travel time is cast into a framework of nonstationary stochastic process. For a fixed value of departure time in a day, travel time given origin, destination, and route information, is treated as a random variable. For a specific date, travel time is treated as a deterministic function of departure time t. Under this framework, the expected travel time for a given departure time is defined as an ensemble mean travel time (EMTT) over a number of days. The method of moment is proposed to compute EMTT based on a hypothetical piecewise constant speed trajectory for travel time estimation. The advantage of the method of moment for EMTT estimation is that it only requires ensemble mean and ensemble variance of spot speed information at point detectors, which is much easier and cost-effective to get than obtaining collections of massive spot speed data per se. The result is compared against Monte Carlo simulation and direct sampling based simulation. The proposed method of moment approach provides accurate estimation of EMTT (e.g., the expected travel time estimation) under a wide range of traffic conditions (e.g., free flow and congestion).
23
Transportation Research Record
Anne Vingaard Olesen, Erik Kjems, Kristian Hegner Reinau et al. · 2016 · 11.4K citations