Dynamic ramp metering strategies designed to actively counter developing bottlenecks can reduce vehicle delay up to 48 percent.
Experience using real-time traffic data to improve ramp metering and mainline performance on Highway-100 in Minneapolis, Minnesota.
Minneapolis, Minnesota, United States
Development of Active Tra ffic Management Strategies for Minnesota Freeway Corridors
Summary Information
A new metering strategy was implemented on Highway-100 Northbound (100NB) in Minneapolis, Minnesota. The strategy activated metering controls for each ramp depending on mainline conditions and the current traffic demand at a given ramp.
To assess system impacts, traffic performance on the mainline and ramp were analyzed before and after the system was implemented. Traffic data were collected during weekday afternoon peak periods (Tuesdays, Wednesdays, and Thursdays). To account for seasonal variations in traffic demand, the performance of the new metering strategy was compared with that of the old strategy during the same months of the previous year.
FINDINGS
Results of the field test indicated substantial improvements in both mainline and ramp traffic performance as compared to the old strategy that used a stratified algorithm. By dynamically configuring the bottleneck-based zone structure to be controlled in real-time, the new strategy did not require pre-specified associations between ramps and potential bottlenecks, thus increasing its flexibility in dealing with incidents or unexpected events. Further, the turn-on/off times of each ramp meter were automatically determined with consideration given to the mainline traffic states. The results below were excerpted from the source report.
Impacts on Travel Delay
Before/After Comparison of Total Delayed Vehicle Hours Traveled
October-November (2011-Before vs 2012-After)
| Before | After | Change | |
| Average | 432.3359 | 224.5156 | -48.1% |
| Variance | 46174.83 | 20186.8 | -56.3% |
| Stddev | 214.8833 | 142.0802 | -33.9% |
Before/After Comparison of Total Delayed Vehicle Hours Traveled
April-May (2012-Before vs 2013-After)
| Before | After | Change | |
| Average | 457.6556 | 377.9004 | -17.4% |
| Variance | 23668.66 | 37924.77 | 60.2% |
| Stddev | 153.8462 | 194.7428 | 26.6% |
Before/After Comparison of Total Delayed Vehicle Hours/Total Entered Vehicles
October-November (2011-Before vs 2012-After)
| Before | After | Change | |
| Average | 0.017575 | 0.008942 | -49.1% |
| Variance | 8.03E-05 | 3.55E-05 | -55.7% |
| Stddev | 0.00896 | 0.005961 | -33.5% |
