Phase II: Operational and Safety-Based Analyses of Varied Toll Lane Configurations

On expressways, managed lanes (MLs) have been introduced as an effective dynamic traffic management strategy. This research consists of two parts: a microsimulation study and driving simulator experiments for appropriate designs for the MLs. The objective of the microsimulation research was to determine optimal access zone density and weaving length. In the simulation, the lane choice replicated drivers’ choice behavior at dynamic tolls based on modeling components and algorithms generated in VISSIM. The network was well calibrated and validated by comparing the operational measurements for simulated and field data. Subsequently, forty-two scenarios were built and tested in VISSIM to specify the optimal accessibility level and to decide on the sufficient weaving distance. The findings indicate that there was a significantly lower conflict risk in MLs than in general-purpose lanes (GPLs). Compared to GPLs, the conflict frequency per vehicle in MLs was less by 48% and 11% in the peak and off-peak traffic conditions, respectively. A Tobit model and a log-linear models were developed for investigating the factors and scenarios that affect traffic conflict frequency. The results of the conflict frequency analysis suggest that one access zone is the optimal accessibility density in the 9-mile segment. Moreover, the results revealed that a length of 1,000 feet per lane change is the optimal length for the weaving segments near access zones. A series of linear regression models was developed to explore the effects of access zone design on the operational performance of the network. The modeling results confirm that one access zone is the optimal level, with a higher speed, a lower delay, and a higher time efficiency than other cases. As the accessibility level increases, the operational performance declines. From the revenue perspective, the case of two access zones creates the largest revenue in the studied network. The traffic operation analysis also revealed that the level of service was the same for the base case with no access zones and the case with one access zone when the weaving distance was higher than 1,000 feet per lane change. The driving simulator experiment aimed to evaluate the impact of different weaving lengths and variable speed limit (VSL) strategy on drivers’ speed control and lane-changing maneuvers. It was found that long weaving lengths (i.e., 1,000 feet and 1,400 feet per lane) resulted in a reduction of average speed and that a weaving length of 1,400 feet per lane had significantly higher speed standard deviation when compared with the other two weaving lengths. In addition, the VSL strategy can reduce the average speed and speed variation. As for the lane-changing behavior, drivers can have the safest performance with 1,000-foot weaving length, in terms of time to collision and number of conflicts. Finally, fewer conflicts could be found in the scenarios with VSL strategy. Another research effort was conducted to compare driving behaviors considering drivers’ gender and age. The experiment results showed that young drivers were prone to drive more aggressively, which resulted in higher speed standard deviation. Also, it was revealed that males have more conflicts when changing lanes than females. It is expected that the results from this study can help engineers and practitioners employ appropriate weaving length, access zone density, and traffic control strategy to enhance traffic operation and safety for MLs.

Language

  • English

Media Info

  • Media Type: Digital/other
  • Features: Figures; Maps; Photos; References; Tables;
  • Pagination: 134p

Subject/Index Terms

Filing Info

  • Accession Number: 01665661
  • Record Type: Publication
  • Files: UTC, TRIS, ATRI, USDOT
  • Created Date: Feb 22 2018 9:32AM