Evaluating the HERO Ramp-Metering Algorithm with San Diego’s Integrated Corridor Management System Model

The FHWA project “Alternative Designs to Alleviate Freeway Bottlenecks at Merge/Diverge and Weaving Areas” aims at evaluating six different methods to mitigate merge impacts. In this paper, the following ramp-metering algorithms were tested: HEuristic Ramp metering coOrdination (HERO), Asservissement Linéaire d’entrée sur Autoroute (ALINEA), and San Diego Ramp Meter System (SDRMS). These were compared with a “do-nothing” base scenario. The algorithms were tested during a 5 h morning peak period simulation using the integrated corridor management system (ICMS) Aimsun network, a complete and wide-ranging network covering San Diego’s I-15 mainline corridor, on-off ramps and arterial roads. Whereas ALINEA is already included in the base distribution of Aimsun, a new implementation of HERO was programmed from algorithms found in the literature and adapted for use on such a network. Different measures of performance (MOP) were used to assess the efficiency of each algorithm and HERO was found to outperform all the other algorithms: gains of 1.5% over ALINEA and 4% over do-nothing on the mainline average travel time, and gains of 0.5% over ALINEA and 1.5% over do-nothing for the average weighted harmonic speed on all mainline and ramp sections. However, although tests showed that better results can be obtained on the mainline, careful calibration was needed to attain overall positive MOP and not penalize vehicles entering on ramps. The paper concludes with proposed improvements to the original algorithm.

Characteristics and Temporal Stability of Recurring Bottlenecks

This study applies and updates a method which identifies and quantifies the extent of traffic congestion from recurring freeway bottlenecks. Additionally, the spatiotemporal stability of bottlenecks over an extended period was tested. Over time congestion at bottlenecks may increase, may decrease, or may migrate to other nearby locations. Analysis of stability is important since prioritizing and applying treatments at bottlenecks is a multiyear process. In addition, a robust method for selecting sensitivity based parameters to identify and quantify bottleneck effects is presented. Subsequently, a systematic framework is developed for tracking and archiving the spatiotemporal changes in the recurring bottlenecks. The proposed method is demonstrated on a case study on Interstate 40 in North Carolina using three years of probe data. A congestion speed ratio detection threshold of 0.7 and a probability of activation threshold of 33% for the study area were determined from a sensitivity test to ascertain their recurrence. The method identified 13 bottlenecks with their impacts ranging from 35 to 3,278 mi-hours of congestion per year. Eight bottlenecks either newly emerged or had their queues merged or shifted between successive years. Even spatially stable bottlenecks had significant variation in their activation frequency and queue length. Exploration of the changes in bottleneck severity and locations revealed the influence of a long-term work zone in the area and the effect of the rapid growth in traffic demand. Local agencies can use this method to shortlist recurring bottlenecks and track changes to plan mitigation strategies.