Variable Speed Limit model yields up to 42.4 percent reduction on number of vehicle stops and 17.6 percent reduction on the average travel time.

Simulation study finds that proactive VSL control models can significantly reduce travel time and the number of vehicle stops over recurrent bottleneck locations.

Date Posted
07/31/2017
Identifier
2017-B01147
TwitterLinkedInFacebook

Proactive Optimal Variable Speed Limit Control for Recurrently Congested Freeway Bottlenecks

Summary Information

This study presents two models for proactive Variable Speed Limit (VSL) control on recurrently congested freeway segments.

  1. BASIC PROACTIVE CONTROL MODEL: a calibrated macroscopic traffic flow model was employed to predict traffic conditions over the target time horizon and to provide the essential information for the VSL optimization module.
  2. ENHANCED PROACTIVE CONTROL MODEL: To fully capture the complex interrelations between speed and density, this study has further adopted a special algorithm to improve the accuracy of the predicted traffic conditions over the target time horizon for proactive VSL control

To evaluate the two proposed optimization models, the study selected the segment MD-100 West from MD 713 to Coca-Cola Drive near BWI Airport. On the selected freeway segment, two VSL signs and four sensors along with a License-plate-recognition (LPR) system were implemented. Both models were investigated using a VISSIM simulator with calibrated field data collected from the Maryland demonstration site Dec. 2009 - Jan. 2010, different traffic conditions and different control objectives.

To compare the proposed VSL models with the No-control scenario, the study designed the following four scenarios:

  • Scenario-1: the basic proactive model with the objective of total travel time minimization;
  • Scenario-2: the enhanced proactive model with the objective of total travel time minimization;
  • Scenario-3: the basic proactive model with the objective of speed variance minimization; and
  • Scenario-4: the enhanced proactive model with the objective of speed variance minimization.

RESULTS

Table Performance Comparison between Different Scenarios

Scenario
6:00-8:00
Ave. # of Stops
Improvement
6:00-8:00
Ave. Travel Time (s)
Improvement
7:00-8:00 Ave. Travel Time (s)
Improvement
No-VSL
3.78
/
215
/
302
/
Scenario-1
4.23
11.9%
240
11.8%
344
13.7%
Scenario-2
3.03
-19.9%
194
-9.6%
270
-11.0%
Scenario-3
3.27
-13.53%
201
-6.2%
280
-7.6%
Scenario-4
2.18
-42.4%
177
-17.6%
235
-22.4%
Goal Areas
Results Type
Deployment Locations