In this presentation we will give an overview of the research done in the Delft Center for Systems and Control in connection with control of freeway traffic networks. In particular, we will focus on model predictive control (MPC) and multi-agent control of freeway traffic networks.
MPC is a model-based predictive control strategy that uses on-line optimization to compute optimal control inputs and settings for the variable speed limit sign, ramp metering installations, and other traffic control measures so as to optimize the travel times of the drivers and/or to reduce the effects of traffic jams and shock waves while at the same time taking various constraints (such as maximum queue lengths, minimum and maximum metering rates or displayed speed limit differences, etc.) into account.
If we consider larger traffic networks, then a centralized control approach is not tractable any more, and a decentralized control approach has to be adopted. In this context, we present a proof-of-concept demo of hierarchical multi-agent traffic control decision support system for the ring-road around Amsterdam.
For more information: click here
Back to Control Seminars Page