Using Archived ITS Data to Measure the Operational Benefits of a System-wide Adaptive Ramp Metering System

Abstract

Type of Report and Period CoveredFinal Report Sponsoring Agency Code Supplementary Notes AbstractA System-Wide Adaptive Ramp Metering (SWARM) system has been implemented in the Portland, Oregon metropolitan area, replacing the previous pre-timed ramp-metering system that had been in operation since 1981. SWARM has been deployed on six major corridors and operates during the morning and afternoon peak hours. This report presents results of a -before‖ and -after‖ evaluation of the performance of two freeway corridors as part of ongoing efforts to measure the benefits of the new SWARM system, as compared to the pre-timed system. The study benefited from using the existing regional data, surveillance and communications infrastructure in addition to a regional data archive system. The evaluation revealed that the operation of the SWARM system, as currently configured in the Portland metropolitan region, produced mixed results when comparing the selected performance metrics to pre-timed operation. For the I-205 corridor, the results were generally positive. In the morning peak period, SWARM operation resulted in an 18.1% decrease in mainline delay and decreased variability in the delay. For the afternoon peak period, improvements were also found (a 7.9 % decrease in mainline delay) with the exception of moderately congested days which saw an 4.7% increase in mainline delay. On the OR-217, however, significant increases were found in overall average delay. In the morning peak period, delay increased 34.9% while in the afternoon period delay increased 55.0%. These conclusions, however, must be tempered because of lack of ramp demand data. If an assumption is made that ramp demand changes correspond with the measured freeway VMT changes, it is likely that ramp delay decreased under SWARM operation (i.e. more vehicles were allowed on the freeway which would equate to lower delay for vehicles on the ramps). Another important finding of this evaluation was that implementation of the SWARM algorithm resulted in significantly more data communication failures in the traffic management system. While this outcome is specific to the ODOT communication infrastructure and hardware, it was not anticipated. These communication failures have the potential to impact other traveler information programs that depend on the freeway surveillance data as well as the SWARM algorithm. Finally, one of the intentions of this research project was to encourage ongoing evaluation and continuous improvement of the ramp metering system and, in general, the overall freeway management system. It is clear from the analysis that meter activation times and rates are necessary to evaluate system performance. Incorporating additional logging capabilities into the SWARM system would make it easier to evaluate system operations on an on-going automated basis.17.

The Use of Cooperative Approach in Ramp Metering

“…It enables an override of the locally-determined metering rates when necessary. A comprehensive evaluation study done in Denver, Colorado in 1988 and 1989, has shown that the central override is effective in mitigating congestion on the mainline if the speeds are less than 90 [km/h] [14].…”

“…These upstream on-ramps are categorized as "slave" on-ramps. The main idea is to exploit their queuing capacity in order to mitigate downstream congestions [14].…”

Improved bottleneck approach for heuristic systemwide responsive ramp metering on congested freeways

“…The Helper method normally cannot achieve the optimization of the entire system. For the SWARM methods, a good prediction model is essential for a successful implementation (Bertini and Hn, 2006), which is still under investigation in this current stage. For the Bottleneck method, considering its advantages such as real-time applicability, limited input parameter, and ease of implementation, it is selected in this paper as the base approach for enhancement.…”

Differential Evolution Approach to Calculate Optimal Ramp Metering Rates

“…These ideas are realized in the SWAR M (Bertini et al, 2008) and the Bottleneck (Lipp et al, 1991;Hochbaum and Woeginger, 1999) algorithms.…”