Feedback control of highway congestion by a fair on-ramp metering

The problem considered in this paper is motivated by the fact that the existing feedback control approaches to highway congestion prevention may result in an unfair capacity allocation whereby one ramp may capture most of the highway capacity, depending on its relative location in the network. To avoid this problem, a new control architecture is proposed. In the framework of this architecture, each segment of the highway is equipped with a congestion controller which, according to a decentralized control law and local measurements, calculates the traffic admission rate beneficial from this segment's point of view. These rates are transmitted to the on-ramps which, according to a control protocol, select the traffic admission rate beneficial from the network's point of view. Thus, the system under consideration is represented by a set of decentralized controllers interacting through the network. It is shown that: (1) the steady states of the system satisfy the so-called max-min fairness property; (2) there exists a set of controller gains that ensure asymptotic stability of these steady states; (3) an adaptive method for the selection of these gains is proposed. Finally results of a numerical investigation, which support the conclusions of the analytical study, are reported.<
>