Ramp metering strategies: a methodology for testing the performance of microsimulation models

During recent decades, traffic demand has overtaken the overall capacity of existing highway and freeway networks, causing traffic delays and negative environmental impacts. Consequently, innovative traffic control strategies have been developed to improve network performance and they have proved to be effective in delaying or preventing congestion. Dynamic microsimulation models are natural contenders for reproducing the impact of traffic control measures on driver behavior and assessing their effect on traffic flows. These models aim to reproduce the flow on an entire network, by modeling individual driver behavior. Many studies have focused on the validation of microscopic models under standard (not controlled) situations. However, empirical evidence shows that traffic control strategies act on standard driver behavior and consequently, the ability of these models to reproduce traffic features under traffic control strategies can be questioned. This paper proposes a comprehensive methodology for measuring the impact of ramp-metering strategies on microscopic simulation results. This methodology relies on traffic flow theory fundamentals and detailed basic traffic scenarios to precisely measure the capacity of metered and unmetered merges. This methodology is easy to implement and can be applied to any microsimulation simulation software. It has been applied to two commercial microsimulation softwares. Surprisingly, (i) the first microsimulation software concludes that ramp-metering strategies can increase or decrease the capacity of a merge, depending on the ramp-metering strategy (one vs several cars per green light); (ii) the second simulation software concludes that ramp-metering strategies have no impact on the capacity of a merge. The results highlight the potential and the limits of microscopic traffic flow models to reproduce complex traffic features involved in traffic control strategies and pave the way for alternative tools to measure the effectiveness of traffic control strategies.