Scenario-based Distributed Model Predictive Control for freeway networks

Abstract

In this paper we develop a scenario-based Distributed Model Predictive
Control (DMPC) approach for large-scale freeway networks. The
uncertainties in a large-scale freeway network are categorized into
global uncertainties for the overall network and local uncertainties for
subnetworks. A reduced scenario tree is proposed, consisting of global
scenarios and a reduced local scenario tree. For handling uncertainties
in the scenario-based DMPC problem, a min-max setting is considered. A
case study is implemented for investigating the scenario-based DMPC
approach, and the results show that in the presence of uncertainties it
is effective in improving the control performance with the queue length
constraint being satisfied.