Optimum Number of Connectors based on the Proximity of Assignment Results to Observed Traffic Counts, Case Study of Mashhad

Traffic volume prediction on links for exact planning and policy making requires the investigation of the impacts of connectors as important elements of the networks, and their optimum number. Few studies have been conducted on the effect of the number of connectors on traffic assignment; however they have emphasized their important role. The aim and novelty of this paper is the prediction of the number of connectors based on proximity of assignment results to observed traffic counts, implemented for case study of Mashhad. Descriptive statistics analysis shows that for the one-connector scenario, the minimum, maximum and standard deviation of the absolute error term are the least among the six scenarios. The mean, however, is not as small as the five-connector scenario, whose other statistics are not as good as the first. Results show that increasing the number of connectors not only does not improve the replication of observed traffic counts, but also causes a decrease in its coefficient of determination (RP2P) from 0.54 for one connector to 0.38 for six connectors for the case study. It is concluded that for a network like that of Mashhad and morning peak demand, the assignment results for only one connector having the best observation replication among the six scenarios.