Multi-class local ramp metering to reduce traffic emissions in freeway systems

Abstract

In this paper a local ramp-metering strategy is proposed to minimize traffic emissions in a freeway; the control scheme is based on a two-class dynamic model in which two vehicle categories related to cars and trucks are explicitly represented. In fact, a major objective of the proposed approach is to devise separate ramp metering control actions for the two classes of vehicles, supposing that separate lanes at the on-ramps and separate traffic lights are present for fast vehicles and slow vehicles. In order to take explicitly into account the traffic emissions of vehicles, we adopt an average-speed model, according to which the emission factor for a given pollutant depends on the average speed during a trip. Considering such a model, a desired value of the traffic density for each class of vehicles is fixed and a properly adapted PI-ALINEA strategy is proposed. The experimental tests show the effectiveness of the proposed approach.