Ramp metering strategies: a literature review

Abstract

Ramp metering (RM) is a control tool that allows keeping the flow in freeways main stream stable (occupancy below critical values), in order to delay the onset of congestion. A proper application guarantees a lower Total Travel time Spent in the network (TTS) (travel time in the main stream and waiting time at on-ramps), to varying degrees depending on the application. The benefits of RM arise from preventing congestion and/or from preventing queue spillback. In this paper, the state of the art is surveyed, ultimately selecting twenty-one relevant documents reporting case studies that compare the performance of two or more RM strategies. These documents are classified by the strategy used and reported gains (reduction in TTS). Regardless of the strategy, the authors recommend applying RM to highly congested freeways with similar TTS among ramps. Bottlenecks should be clearly identified to establish the place where the control must be carried out. Control parameters should be set using historical and real-time data and updated according to a receding horizon. Detrimental factors that could exacerbate initial conditions of congestion and spillback include: incorrect input parameters, reactive instead of proactive approaches, and queue formation only on some coordinated ramps.