Browsing by Subject "AIMSUN"

Now showing 1 - 1 of 1
  • Thumbnail Image
    Item
    Microscopic Simulation and Evaluation of the Roundabout Capacity Model in Highway Capacity Manual
    (2018-01) Chen, Rongsheng
    Shown to be an effective intersection design, the roundabout is receiving increasing attention and popularity. Several models, described in this work, have been developed to predict roundabout capacity. One of them, the roundabout capacity model included in the Highway Capacity Manual (HCM), is widely used in the US, using a gap-acceptance foundation based on data collected in US roundabouts. This study explored the accuracy of the two-lane variants of the roundabout capacity models in HCM 6th Edition and HCM 2010 by comparing them with an exponential regression model fitted on flow rate measurements collected at a two-lane roundabout in Richfield, Minnesota. Based on the same gap-acceptance foundation proposed in HCM, two other models were developed by recalculating coefficients. Each followed a different calibration strategy and compared with the Richfield model. It was found that calibration can significantly enhance the accuracy of the default HCM model and calibrating only the intercept of the default HCM model can produce a model with similar accuracy as the model resulting by calibrating both coefficients. To further assist traffic engineers, this work validated the capability of the popular traffic simulator AIMSUN to build a roundabout model with realistic capacities. A sensitivity analysis, exploring the impact of different simulation parameters, further assisted in proposing an efficient and reliable simulation calibration methodology. Initial safety margin, visibility along main stream, reaction time at stop, and max acceleration were selected to calibrate driver’s gap acceptance behavior. The result showed that if a calibrated model in AIMSUN could produce the same critical headway and follow-up headway as those in the HCM6 model, it will also result in similar capacities as the HCM6 model.