Data-driven generalized perimeter control: Zürich case study

Abstract

Urban traffic congestion is a key challenge for the development of modern cities, requiring advanced control techniques tooptimize existing infrastructures usage. Despite the extensive availability of data, modeling such complex systems remainsan expensive and time consuming step when designing model-based control approaches. On the other hand, machine learningapproaches require simulations to bootstrap models, or are unable to deal with the sparse nature of traffic data and enforce hardconstraints. We propose a novel formulation of traffic dynamics based on behavioral systems theory and apply data-enabledpredictive control to steer traffic dynamics via dynamic traffic light control. A high-fidelity simulation of the city of Zürich, thelargest closed-loop microscopic simulation of urban traffic in the literature to the best of our knowledge, is used to validate theperformance of the proposed method in terms of total travel time and CO2 emissions.