Multi-agent model predictive control for transportation networks with continuous and discrete elements

Multi-agent model predictive control for transportation networks with continuous and discrete elements

Negenborn, R.R.
De Schutter, B.
Hellendoorn, J.

Mechanical, Maritime and Materials Engineering

Delft Center for Systems and Control

2006-08-01

For several reasons, control of transportation networks like road traffic networks, power transmission networks, water distribution networks, etc. has to be done using a multi-agent control approach. We consider a multi-agent control approach in which each control agent uses Model Predictive Control (MPC) to determine its actions. Each control agent uses a model of its subnetwork and communication with its neighboring control agents to come to agreement on the evolution of interconnections and to determine optimal local inputs and states. The challenge that is addressed in this paper comes from the situation that arises when the models that the agents use contain both continuous and discrete elements, instead of one of the two exclusively. We propose an approach that deals with this hybrid behavior and results in agreement between the agents with respect to the evolution of interconnections and that moreover provides integer inputs to obtain these evolutions.

multi-agent systems
model predictive control
transportation networks
hybrid systems

http://resolver.tudelft.nl/uuid:8a99b1ca-6ea6-4578-9bb0-a843150d5c39

International Federation of Automatic Control (IFAC)

Proceedings of the 11th IFAC Symposium on Control Transportation Systems, IFAC-CTS 2006, 29-31 August 2006, Delft, The Netherlands

Institutional Repository

conference paper

(c) 2006 The Author(s)
IFAC