Advanced control techniques for optimal adaptive traffic control

Project members: A. Hegyi, B. De Schutter, J. Hellendoorn

Sponsored by: AVV (Transport Research Centre, Ministry of Transport, Public Works and Water Management, The Netherlands)

The overall framework of this project is dynamic traffic management (DTM). We mainly address systems and control issues of DTM. More specifically, we investigate the possibilities and advantages of using advanced control techniques in optimal adaptive traffic control.

Traffic patterns change during the day and depend on external influences such as weather conditions, incidents, holidays, and so on. In order to obtain optimality, traffic control policies should adapt to these changes. Adaptive controllers take the changes in the traffic system and the external conditions into account and in that way they can deal with the changes in the traffic patterns. Therefore, we consider model predictive control (MPC) traffic controllers in this project.

In this project we concentrate on two traffic control measures for motorway traffic: ramp metering (see also Project 7.2 and Figure 25) and variable speed limits [25,24]. More specifically, we apply MPC to optimally coordinate variable speed limits and ramp metering for highway traffic flow control. The basic idea is that speed limits can increase the (density) range in which ramp metering is useful. For the prediction we use a slightly adapted version of the METANET traffic flow model that takes the variable speed limits into account. The optimal control signals aim at minimizing the total time that vehicles spend in the network. The coordinated control results in a network with less congestion, a higher outflow, and a lower total time spent. In addition, the receding horizon approach of model predictive control results in an adaptive, on-line control strategy that can take changes in the system automatically into account.

Figure 25: Ramp metering on the A13 highway in Delft-Zuid.