Two system architectures for traffic management are proposed in this paper. Both are based on the concept of decision support system, and propose decision support systems for traffic management for two types of situations: a road network managed by a unique traffic control centre, and an area, possibly a metropolitan one, subdivided in zones each one under the control of a different traffic control centre with possibly different control strategies, that would require a certain level of co-ordination to improve the performance of the whole system. The proposed architectures are based on a short term demand forecasting and a microscopic simulator to estimate network states, current and forecast. The paper mainly discusses the role of the microscopic simulation, its computing performance for real-time applications when parallelisation is taken into account, and briefly discusses some preliminary results on the system validation. (A)
Abstract