Adaptive signal control integrated with freeway ramp meters in Glasgow, Scotland improved network travel times by 10 percent.
Date Posted
09/10/2002
Identifier
2007-B00337
TwitterLinkedInFacebook

Application and Evaluation of the Integrated Traffic-Responsive Urban Corridor Control Strategy (IN-TUC) in Glasgow

Summary Information

As part of the European TABASCO (Telematics Applications in Bavaria, Scotland, and Others) project, traffic-responsive urban control (TUC) strategies were developed and applied to the M-8 corridor in Glasgow, Scotland. Between 1997 and 1998 TUC strategies were deployed. Aaptive signal control, highway ramp metering, and dynamic message signs (DMSs) were used to balance network traffic loads on arterials and freeways during periods of heavy congestion.

  • Traffic signal timing at intersections were optimized every 2 minutes using linear-quadratic programming to balance the number of vehicles traveling on urban links.
  • Ramp meter wait times were adjusted every 60 seconds using the ALINEA (Asservissment LINeaire d’Entrée Autoroutière) strategy to maintain freeway flow downstream of merging locations.
  • DMS boards were used to provide alternative route information and equalize travel times between competitive routes.
The evaluation project considered the impacts of TUC on the entire urban network as well as areas in and around the deployment area. The study area included 44 links. Thirty-nine (39) of these were urban links, three were freeway links, and two were freeway on-ramp links (one ramp was metered). Seven (7) of these links had signalized junctions, and three links had DMSs. The evaluation was completed in March of 1998.


SIMULATION RESULTS

The simulated network covered 3.5 kilometers (km) of the M8 corridor and adjacent arterials, and included 27 intersections/junctions with seven intersections manipulated by the TUC adaptive signal control strategy. The METACOR (Modèle d’Ecoulement du TrAfic sur CORidor) macroscopic simulation model was used to model the following five scenarios. using a representative four hour time horizon:

  • Low demand conditions.
  • High demand conditions.
  • Demand fluctuations.
  • Occurrence of incident before peak conditions.
  • Occurrence of incident during peak conditions.

A basic store-and-forward modeling approach was used to push vehicles onto the network based on demand conditions for each scenario. Vehicles on identical links experienced constant travel times, however, if the link inflow was greater than the link outflow no additional vehicles were added to the link until space became available. The outflows at intersections and junctions with adaptive signal control systems were updated periodically to optimize network traffic flows.

The METACOR simulation model was validated against field data during the TABASCO project and found to realistically reproduce traffic conditions in the Glasgow network.

The simulation study indicated that if TUC adaptive signal control strategies were applied to 7 out of 27 fixed time signalized intersections and junctions, vehicle delay within the network would decrease 1 to 2 percent, and vehicle delay at intersections manipulated by TUC would decrease 5 to 7 percent.

FIELD STUDY RESULTS

In July 1997, ramp metering was deployed, and by February 1998 the other components (urban intersection traffic control (UITC), and DMSs) became fully operational. The project evaluation was completed in March 1998.

The impacts of the system on journey times were measured for several weekdays during afternoon peak periods of congestion. The data presented in the table below was compared to baseline conditions and shows the impacts of ramp metering alone, the impacts of ramp metering and UITC, and the impacts of ramp metering, UITC, and DMSs).


Effect of Control on Journey Time (PM Peak 16:00-17:00)

Base Case (Seconds)
Change with Ramp Metering
Change with Ramp Metering and UITC
Change with Ramp Metering, UITC, and DMSs
Motorway
210
–5%
*
*
Urban Diversion Routes
440
+4%
*
+1%
Total Urban Network
1174
*
–11%
–15%
Total Evaluation Network (including the metered on-ramps)
1567
–2%
–10%
–13%

* Not statistically significant
UITC = Urban Intersection Traffic Control
DMSs = Dynamic Message Signs

Field testing results were compared to simulation results. The field test study area included 44 links. Thirty-nine (39) of these were urban links, three were freeway links, and two were freeway on-ramp links (one ramp was metered). Seven (7) of these links had signalized junctions, and three links had DMSs.
Goal Areas
Deployment Locations