Short-Term Vehicle Arrival Distribution Verification of Expressway Toll Stations Based on ETC Traffic Flow

During traffic peak hours, the manual toll collection system (MTC) toll lanes of the expressway are prone to congested queues. Most of the references directly use the queuing theory model optimizing the MTC toll lane configuration. Usually, they assume the number of vehicles arriving in a short period of time satisfies a Poisson distribution, however, few studies have tested the validity of this hypothesis. By using the electronic toll collection (ETC) traffic flow data to perform a K-S non-parametric test on the vehicle arrival distribution of the toll station in a short time, this paper verifies that arrival of vehicles within 10 consecutive minutes obeys the Poisson distribution. Moreover, by using a k-s nonparametric test, this paper finds that the mean value of the independent Poisson sample is asymptotic normal distribution.