| A set of data, collected from a ramp-weave area, is analyzed to investigate the behavior of traffic in freeway weaving areas. The data were developed by digitizing the trajectory of every vehicle passing through the weaving section at one-second intervals during one-hour period. The main emphasis of this study is placed on analyzing the lane-changing pattern and gap acceptance characteristics of weaving vehicles. Descriptive statistics of the lane-changing point and the accepted lead, lag, and gap are studied to investigate the general trends in the weaving behavior. Also, the multiple linear regression technique is utilized to explore the relationships between these variables and various factors including traffic condition parameters. Major findings of this study include: (a) the behavior of weaving traffic is more strongly associated with traffic density and speed than with traffic volume, (b) traffic conditions in the origin and destination lanes appear to be more effective than the average traffic conditions in the freeway overall for explaining the behavior of weaving vehicles, and (c) late merging vehicles do not necessarily accept smaller gaps than do early merging vehicles. |
