Model of Capacity Estimation of Roundabout in Beijing

2014 
Weaving sections have been shown to have a great effect on capacity of a roundabout, because at weaving sections the traffic conflict and interference are predominant, resulting in a reduction of speed as well as increase of gaps. It is known that the weaving section of a roundabout is the bottleneck of the traffic movement. Therefore, it should be given the highest priority in the capacity analysis. The current methods of estimating the approach capacity of a roundabout mainly utilize information of vehicles entering and exiting the roundabout. From a literature review it has been found that few studies are conducted based on the traffic characteristics of a weaving section in capacity estimation. The purpose of this research is to estimate the capacity of a roundabout by modeling the radius of the roundabout, width of circulating roadway, and gap acceptance at the weaving sections. For this reason, a total of 21 roundabouts that are located at the metropolitan Beijing are selected to conduct data collection. Significance analysis of speed at weaving sections has been conducted by means of the variance analytic method. By taking geometric conditions and traffic conditions into account, new concepts of “Quasi-saturation state of roundabout” and “The Capacity of Roundabout” are introduced in this study. The regression models of the headway and the radius on both the outer lanes and the inner lanes are established respectively using Statistical Program for Social Sciences (SPSS 18.0) for Windows. Moreover, based on the theory of saturation flow rate, a new method is proposed to estimate the capacity of single-lane and multi-lane roundabouts using MATLAB software. And then the real data are taken to validate this model. The findings from this research demonstrate that capacity estimates with weaving section vehicles result in improved prediction of the actual capacity of a roundabout.
    • Correction
    • Source
    • Cite
    • Save
    • Machine Reading By IdeaReader
    11
    References
    0
    Citations
    NaN
    KQI
    []