Strategic control in metropolitan areas [road traffic]

Traffic adaptive control systems proved to perform better compared to fixed time or traffic actuated control systems. In particular adaptive network control can take into account more global objectives, e.g. minimisation of delay or of travel time within the considered system. For this reason more and more adaptive control methods are being implemented in metropolitan areas. They include traffic signal control i.e. variable direction signs, ramp control, dynamic lane assignment and parking guidance. The methods are operating responsively to stochastic variations and to the change of traffic demand over time according to a given objective function. However, the different systems within a city are in many cases not integrated by a higher level and thus may aim to different objectives. Moreover, there may be the need for common policies which contradict the once given objectives of the respective control systems. This paper introduces the idea of a distributed but hierarchically structured system architecture that provides a strategic level by which various different systems can be integrated. The paper further focuses on the methodology by which traffic policies may be translated into the objective functions of single control systems. To show the relevance of strategic control for the day to day operation, scenarios for real applications are given for the example of Munich