خلاصة:
Traffic flow characteristics depend on driver behavior, vehicle specifications, physical characteristics of the route, and the interaction of these factors with each other. In such a way that even if the passing vehicles and geometric characteristics of the routes are similar, the traffic flow characteristics will not necessarily be the same due to differences in the behavior of passing drivers. Although various models have been proposed since 1934 AD to predict traffic flow characteristics, the most appropriate model expressing the relationship between different flow variables remains controversial, mainly due to the difference in the behavior of passing drivers given the existing conditions on the route. In particular, the high dispersion of observations due to their large number makes determining specific points on the curves describing flow variables difficult. This article analyzes traffic flow behavior based on local data, emphasizing the speed-density model for the Tehran-Qom freeway. In the case study of this research, more than four million traffic count data in the Tehran-Qom freeway in two directions (outbound and inbound) and separated by passing lane have been examined and analyzed. Based on these local data, various speed-density models have been presented, evaluated, and analyzed, and their results have been compared with classic past models. The results of this research showed that the weighted average slope of speed changes at densities less than 7 equivalent vehicles per kilometer of the passing lane is approximately 30 percent less than the weighted average slope of speed changes at densities greater than this amount. Finally, comparing the models developed in this research with other types of classic past models indicated higher accuracy of the developed models. Therefore, the models presented in this research provide a better description of the traffic conditions on the freeway given the behavior of local drivers compared to other models.
ملخص الجهاز:
In any case, by obtaining the appropriate modeling pattern (two-regime linear) and clearly determining the density of the regime change point (equivalent to 7 vehicles per kilometer per lane), in the next step, in order to obtain the speed-density model for all observations, modeling was performed using all data, and the regression deviation error of these models was compared with other models (for example, the Greenshields model).
(Refer to the page image) (Refer to the page image) 3- Fitting all traffic data to multi-regime models: As mentioned in sections 5-1 and 5-2, due to the high dispersion of field data on the Tehran - Qom freeway (Charts 7 and 8), if all traffic information is used (without classification into different groups), selecting a single-regime model cannot accurately represent the flow pattern.
It should be noted that the low values of R2 coefficients in the models constructed for densities less than 7 passenger car equivalents per kilometer per lane is one of the results / observations of the current article, which indicates that this issue (high dispersion of observations) is a natural matter due to the greater freedom of action of drivers in choosing speed during free-flow traffic conditions.
In this regard, the low values of R2 coefficients in the constructed models in the state of density less than 7 passenger car equivalents per kilometer per lane is one of the results/observations of the current article, which is due to the greater freedom of drivers in choosing speed in the free-flow traffic state as well as the short length of the transition zone.
(
, 
, 
)
