Path-based joint demand management and route guidance for multi-region traffic networks

This work investigates the joint optimization problem of demand management and route guidance to minimize the time spent of vehicles in traffic networks characterized by well-defined regional fundamental diagrams. To account for variable route length, a path-based regional traffic flow model is introduced which enables the routing of vehicles through multiple paths. Within this modeling framework, route guidance aims to reduce the density imbalances observed within each region by splitting the demand flows of arbitrary origin-destination pairs among a set of preselected paths. At the same time, demand management regulates access to the network by delaying the journey starting time of individual vehicles. For the solution of the considered problem, a Model Predictive Approach (MPC) is proposed which aims towards minimizing the total time that vehicles spend in the network while maintaining non-congested conditions at all times through demand management. A relaxed linear programming formulation of the original non-convex, non-linear problem is also proposed which provides tight lower bound to the optimal solution.